AxonHub: baixar versions anteriors

What's Changed
- feat: unify reasoning effort across protocols and add model endpoint routing by @AkaChou in https://github.com/looplj/axonhub/pull/2330
- fix: dialog double scrollbar introduced in 2bfea194 by @BLumia in https://github.com/looplj/axonhub/pull/2355
- feat: show codex reset details, close #2301 by @looplj in https://github.com/looplj/axonhub/pull/2360
- fix(frontend): repair height-constrained scroll layout in mobile dialogs by @djdembeck in https://github.com/looplj/axonhub/pull/2354
- fix(codex): strip user field from outbound requests by @llc1123 in https://github.com/looplj/axonhub/pull/2352
- fix: preserve transformed Responses stream events by @MrLiuGangQiang in https://github.com/looplj/axonhub/pull/2365
- fix: dockerfile compitable, close #2367 by @looplj in https://github.com/looplj/axonhub/pull/2368
- fix(anthropic): reject invalid tool input on clean EOF by @llc1123 in https://github.com/looplj/axonhub/pull/2363
- feat: add opencode session header, close #2361 by @looplj in https://github.com/looplj/axonhub/pull/2369
Full Changelog: https://github.com/looplj/axonhub/compare/v1.0.0-beta8...v1.0.0-beta9

What's Changed
- opt: anthropic signature recognization by @looplj in https://github.com/looplj/axonhub/pull/2197
- feat: unify auto-refresh controls and stabilize list animations by @AkaChou in https://github.com/looplj/axonhub/pull/2198
- fix: correct currency code example from RMB to CNY by @hanuxxx in https://github.com/looplj/axonhub/pull/2202
- feat(quota): add Charm Hyper credit balance checker by @djdembeck in https://github.com/looplj/axonhub/pull/2199
- fix(quota): color Codex quota bars by usage by @djdembeck in https://github.com/looplj/axonhub/pull/2166
- feat(quota): track OpenCode Go quota via official usage API by @jveko in https://github.com/looplj/axonhub/pull/2204
- feat: hide unroutable configured models from public lists by @CberYellowstone in https://github.com/looplj/axonhub/pull/2215
- fix(trace): only persist explicitly identified traces by @MrLiuGangQiang in https://github.com/looplj/axonhub/pull/2208
- fix(responses): harden stream terminal, retry, and resource boundaries by @wtj-0527 in https://github.com/looplj/axonhub/pull/2196
- feat: add xai_responses channel type by @kirafishy in https://github.com/looplj/axonhub/pull/2212
- fix(prompts): parse project GUIDs in prompt schema by @Likhixang in https://github.com/looplj/axonhub/pull/2200
- chore(actions): upgrade workflows to node24 actions by @llc1123 in https://github.com/looplj/axonhub/pull/2219
- fix(responses): preserve image tool_choice without pass-through by @llc1123 in https://github.com/looplj/axonhub/pull/2218
- fix(codex): skip pass-through for unsupported response limits by @llc1123 in https://github.com/looplj/axonhub/pull/2217
- feat(requests): add reusable column reordering with drag-and-drop by @djdembeck in https://github.com/looplj/axonhub/pull/2222
- feat: persit chunks for failed stream request by @looplj in https://github.com/looplj/axonhub/pull/2224
- feat(quota): allow exempting channels from quota filtering by @djdembeck in https://github.com/looplj/axonhub/pull/2214
- feat: add xAI subscription SSO channel and Responses support for xai by @llc1123 in https://github.com/looplj/axonhub/pull/2225
- feat: provider quota estimate by @looplj in https://github.com/looplj/axonhub/pull/2228
- feat: add codex missing headers by @looplj in https://github.com/looplj/axonhub/pull/2230
- fix(actions): derive unstable version from latest tag by @llc1123 in https://github.com/looplj/axonhub/pull/2231
- fix(llm): preserve top-level reasoning tokens by @djdembeck in https://github.com/looplj/axonhub/pull/2237
- fix(nanogpt): escalate quota status to exhausted at 100% by @djdembeck in https://github.com/looplj/axonhub/pull/2236
- fix(codex): fix HTTP 400 missing reasoning.context all_turns by @djdembeck in https://github.com/looplj/axonhub/pull/2235
- feat(frontend): add sans/serif/mono font options to theme menu by @ChancenJ in https://github.com/looplj/axonhub/pull/2234
- chore: minor ui/error optimization by @looplj in https://github.com/looplj/axonhub/pull/2238
- fix: image outbound & override by @looplj in https://github.com/looplj/axonhub/pull/2241
- fix: restore meta muse lineup in model catalog by @djdembeck in https://github.com/looplj/axonhub/pull/2258
- fix(channel): write UTC updated_at when saving model prices by @ChancenJ in https://github.com/looplj/axonhub/pull/2257
- fix(frontend): grant system-level route access to owner users by @qinkangdeid in https://github.com/looplj/axonhub/pull/2254
- fix(copilot): send Grok models through Responses API by @ripenedcat in https://github.com/looplj/axonhub/pull/2250
- opt: gemini channel compatible by @looplj in https://github.com/looplj/axonhub/pull/2259
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/2242
- feat: introduce unified opencode transformer, close #2240 by @looplj in https://github.com/looplj/axonhub/pull/2260
- feat: GC 支持剥离已存储的请求/响应载荷 by @CberYellowstone in https://github.com/looplj/axonhub/pull/2246
- fix(frontend): 

What's Changed
- fix: handle moderation responses in empty response detection by @Silentely in https://github.com/looplj/axonhub/pull/2099
- feat: add provider quota collection controls by @Silentely in https://github.com/looplj/axonhub/pull/2103
- fix(gemini): preserve file data MIME type for Vertex by @llc1123 in https://github.com/looplj/axonhub/pull/2091
- feat: backup & restore systems config, close #2101 by @looplj in https://github.com/looplj/axonhub/pull/2104
- feat: response with trace headers, close #2100 by @looplj in https://github.com/looplj/axonhub/pull/2107
- feat: check new beta version, close #2097 by @looplj in https://github.com/looplj/axonhub/pull/2111
- fix(frontend): improve requests table mobile responsiveness by @djdembeck in https://github.com/looplj/axonhub/pull/2118
- fix: reuse bundled Windows binary by @Jonesxq in https://github.com/looplj/axonhub/pull/2119
- fix(analytics): support MySQL for channel stats by @ChancenJ in https://github.com/looplj/axonhub/pull/2123
- fix: 修正拼写错误“oltphttp”->“otlphttp” by @bookshiyi in https://github.com/looplj/axonhub/pull/2121
- fix(frontend): load theme fonts by @llc1123 in https://github.com/looplj/axonhub/pull/2122
- fix(responses): keep images in tool results instead of dropping them by @GhostFlying in https://github.com/looplj/axonhub/pull/2117
- feat: add configurable max_multipart_memory for backup restore by @vincentyzhj in https://github.com/looplj/axonhub/pull/2126
- feat: add per-channel API key rule actions by @Lxiny-zy in https://github.com/looplj/axonhub/pull/2120
- fix(transformer): 合并连续 function_call/custom_tool_call 为单条 assistant 消息 by @Snowy117 in https://github.com/looplj/axonhub/pull/2114
- fix(responses): scope fallback prompt_cache_key per conversation by @Kizunad in https://github.com/looplj/axonhub/pull/2031
- feat(permissions): align project roles with UI access by @brynne8 in https://github.com/looplj/axonhub/pull/2095
- feat: model load balance strategy by @looplj in https://github.com/looplj/axonhub/pull/2127
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/2131
- fix: admin page permission protect by @hen7777777 in https://github.com/looplj/axonhub/pull/2138
- fix(responses): handle reasoning text stream events by @cesaryuan in https://github.com/looplj/axonhub/pull/2151
- fix: harden MemorySize parsing and document max_multipart_memory by @vincentyzhj in https://github.com/looplj/axonhub/pull/2145
- fix(openai): keep message content valid for strict upstreams by @Lxiny-zy in https://github.com/looplj/axonhub/pull/2137
- fix(frontend): constrain quota popover width on mobile by @KremeCN in https://github.com/looplj/axonhub/pull/2143
- fix(frontend): prevent system settings double scroll by @KremeCN in https://github.com/looplj/axonhub/pull/2142
- fix(cline): align quota windows with API behavior by @KremeCN in https://github.com/looplj/axonhub/pull/2141
- feat(cline): fetch recommended models dynamically by @KremeCN in https://github.com/looplj/axonhub/pull/2140
- feat(channels): add Groq channel by @rupaut98 in https://github.com/looplj/axonhub/pull/2144
- fix(transform): add channel switch to downgrade mid-conversation system messages by @AkaChou in https://github.com/looplj/axonhub/pull/2124
- feat(requests): show final reasoning_effort in request logs by @AkaChou in https://github.com/looplj/axonhub/pull/2158
- feat(channels): unified API key management dialog with bulk import by @Lxiny-zy in https://github.com/looplj/axonhub/pull/2156
- fix(cline): normalize empty message content by @KremeCN in https://github.com/looplj/axonhub/pull/2155
- fix(frontend): align analytics filter UX by @cesaryuan in https://github.com/looplj/axonhub/pull/2154
- fix(responses): preserve function calls completed in done events by @MrLiuGangQiang in https://github.com/looplj/axonhub/pull/2057
- feat(requests): redesign request log table by @brynne8 in https://github.com/looplj/axonhub/pull/2162
- perf(

What's Changed
- fix(dashboard): fix daily chart X-axis date display using Date.UTC by @ChancenJ in https://github.com/looplj/axonhub/pull/1995
- fix: dialog ui for test channel when enable channel, close #1969 by @looplj in https://github.com/looplj/axonhub/pull/1996
- fix(anthropic): synthesize terminal stream events when upstream omits finish_reason by @mvanhorn in https://github.com/looplj/axonhub/pull/2000
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/2004
- feat: display last name before first name in sidebar for Chinese locale by @hihanley in https://github.com/looplj/axonhub/pull/1998
- fix(apikeys): avoid double-counting cached and reasoning tokens in total by @hihanley in https://github.com/looplj/axonhub/pull/1999
- feat: add IBM model developer with Granite models by @djdembeck in https://github.com/looplj/axonhub/pull/2003
- feat: add analytics page with multi-dimensional filtering by @ChancenJ in https://github.com/looplj/axonhub/pull/2005
- feat(traces): add span statistics to trace detail page and filter pop… by @ChancenJ in https://github.com/looplj/axonhub/pull/2008
- fix: move analytics helpers out of resolvers file to prevent go generate from breaking them by @ChancenJ in https://github.com/looplj/axonhub/pull/2019
- feat: support set-if-absent body overrides by @GhostFlying in https://github.com/looplj/axonhub/pull/2017
- feat: expose prompt cache key to override templates by @GhostFlying in https://github.com/looplj/axonhub/pull/2016
- fix(provider-quota): use official Cline quota limits by @KremeCN in https://github.com/looplj/axonhub/pull/2012
- feat(channel): auto-configure model prices by @ousing9 in https://github.com/looplj/axonhub/pull/1993
- feat: add Kimi Code quota tracking by @u33e in https://github.com/looplj/axonhub/pull/2010
- fix(codex): preserve Responses Lite request requirements by @shay-wong in https://github.com/looplj/axonhub/pull/2020
- fix(traces): normalize tool_use arguments JSON to fix span dedup by @ChancenJ in https://github.com/looplj/axonhub/pull/2013
- feat: config auto reload by @looplj in https://github.com/looplj/axonhub/pull/1997
- fix: embed IANA timezone database for Windows compatibility by @ChancenJ in https://github.com/looplj/axonhub/pull/2024
- feat: add trace and thread status management (archive / retain) by @ChancenJ in https://github.com/looplj/axonhub/pull/2023
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/2026
- feat(quota): add Minimax channel quota checker by @ChancenJ in https://github.com/looplj/axonhub/pull/2029
- fix(openai): write image edit model before files by @llc1123 in https://github.com/looplj/axonhub/pull/2028
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/2037
- feat(quota): add Zhipu channel quota checker by @ChancenJ in https://github.com/looplj/axonhub/pull/2032
- fix(ent): align prompt project ownership by @brynne8 in https://github.com/looplj/axonhub/pull/2039
- feat(models): add Tencent developer with Hunyuan models by @KtzeAbyss in https://github.com/looplj/axonhub/pull/2046
- feat: add archive/retain buttons to trace and thread detail pages by @ChancenJ in https://github.com/looplj/axonhub/pull/2045
- fix(channels): preserve third-party Codex base URL by @brynne8 in https://github.com/looplj/axonhub/pull/2040
- feat(models): add poolside as model developer by @djdembeck in https://github.com/looplj/axonhub/pull/2059
- feat(channels): add Anthropic messages API format support to the Ollama channel by @hihanley in https://github.com/looplj/axonhub/pull/2058
- feat: 支持配置全局渠道测试 Prompt by @CberYellowstone in https://github.com/looplj/axonhub/pull/2042
- feat:增加每次请求新连接功能，使得类似 resin 之类的代理池能够轮换 ip by @xuyufengfei in https://github.com/looplj/axonhub/pull/2060
- fix(permissions): align system settings scope by @brynne8 in https://github.com/looplj/axonhub/pull/2053
- feat: add ti

What's Changed
- opt(transformer): build image data URLs via xurl.BuildDataURL by @suixinio in https://github.com/looplj/axonhub/pull/1882
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1887
- fix: passThrough column not toggleable in column visibility settings by @Silentely in https://github.com/looplj/axonhub/pull/1886
- fix:跨渠道重试成功之后将原失败渠道记入渠道面板的监控统计 by @xuyufengfei in https://github.com/looplj/axonhub/pull/1891
- fix(openai): avoid stream aggregation nil panics by @luoxue03 in https://github.com/looplj/axonhub/pull/1890
- feat(openapi): support querying api keys and templates by name by @suixinio in https://github.com/looplj/axonhub/pull/1838
- fix: tracing document config by @looplj in https://github.com/looplj/axonhub/pull/1897
- opt: admin page permission guard, close #1894 by @looplj in https://github.com/looplj/axonhub/pull/1898
- fix(frontend): avoid logout on upstream auth errors by @lihaojiee in https://github.com/looplj/axonhub/pull/1902
- fix(responses): preserve encrypted reasoning item boundaries by @ttttmr in https://github.com/looplj/axonhub/pull/1901
- fix: test request detail nav back, close #1912 by @looplj in https://github.com/looplj/axonhub/pull/1915
- feat: add scripts to cleanup sqlite, close #1905 by @looplj in https://github.com/looplj/axonhub/pull/1916
- opt: auto set default test model, close #1895 by @looplj in https://github.com/looplj/axonhub/pull/1917
- fix(ollama): extract token usage from final stream chunk for correct status by @myml in https://github.com/looplj/axonhub/pull/1923
- feat(orchestrator): enforce strict per-channel max concurrency by @suixinio in https://github.com/looplj/axonhub/pull/1920
- fix: scope-level aware RouteGuard and silent 403 handling by @brynne8 in https://github.com/looplj/axonhub/pull/1909
- fix(playground): only show thinking in last assistant message by @myml in https://github.com/looplj/axonhub/pull/1461
- fix: clear the previous state, close #1918 by @looplj in https://github.com/looplj/axonhub/pull/1925
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1934
- feat: add OpenCode Go provider quota tracking by @zhaozhaozz in https://github.com/looplj/axonhub/pull/1926
- fix(channels): fix test API keys dialog footer buttons being hidden by @Silentely in https://github.com/looplj/axonhub/pull/1942
- fix(channels): Update the Base URL of the Ollama service provider to the official address. by @Silentely in https://github.com/looplj/axonhub/pull/1941
- fix: improve error reporting for /v1/models endpoint by @trim21 in https://github.com/looplj/axonhub/pull/1931
- opt: reduce S3 Class A operations with native ObjectStore by @suixinio in https://github.com/looplj/axonhub/pull/1939
- fix(channel): add missing optional() to modelMappings zod schema by @junjiangao in https://github.com/looplj/axonhub/pull/1946
- fix: retry stream errors before first content by @lihaojiee in https://github.com/looplj/axonhub/pull/1944
- feat: add round-robin load balancer strategy by @yuanhhs in https://github.com/looplj/axonhub/pull/1953
- feat: add personal API key type and per-user usage statistics by @brynne8 in https://github.com/looplj/axonhub/pull/1933
- fix(httpclient): cap upstream error response body at 1 MB to prevent OOM by @djdembeck in https://github.com/looplj/axonhub/pull/1955
- feat: ip access control by @looplj in https://github.com/looplj/axonhub/pull/1956
- chore: remove the test source filter for request, close #1913 by @looplj in https://github.com/looplj/axonhub/pull/1957
- feat: show reasoning token usage by @looplj in https://github.com/looplj/axonhub/pull/1960
- feat: add codex headers by @looplj in https://github.com/looplj/axonhub/pull/1963
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1964
- ci: publish Helm chart to GHCR --issue=#1965 by @zouxingyuks in https://github.com/looplj/axonhub/pull/196

What's Changed
- fix: preserve filter state when navigating to request detail page by @MoshiCoCo in https://github.com/looplj/axonhub/pull/1819
- opt: add empty tolerance for claude code read, close #1681 by @looplj in https://github.com/looplj/axonhub/pull/1820
- 修复：复制渠道时同步复制模型价格 by @luoxue03 in https://github.com/looplj/axonhub/pull/1811
- opt: gc & retry by @looplj in https://github.com/looplj/axonhub/pull/1823
- fix(i18n): add missing translation keys for 3 anthropic channel variants by @MoshiCoCo in https://github.com/looplj/axonhub/pull/1824
- chore: upgrade claude code oauth by @looplj in https://github.com/looplj/axonhub/pull/1827
- fix(frontend): 允许 Codex 渠道切换协议 by @llc1123 in https://github.com/looplj/axonhub/pull/1808
- feat:在渠道配置中新增“重试错误信息”规则 by @xuyufengfei in https://github.com/looplj/axonhub/pull/1828
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1830
- feat: update default models for bailian, xiaomi, volcengine, zhipu channels by @MoshiCoCo in https://github.com/looplj/axonhub/pull/1829
- feat: codex image gen/edit, close #1816 by @looplj in https://github.com/looplj/axonhub/pull/1831
- chore: channel action dialog minor ui by @looplj in https://github.com/looplj/axonhub/pull/1832
- docs: correct dev Go version floor to 1.26 to match go.mod by @Dashsoap in https://github.com/looplj/axonhub/pull/1833
- opt: openai response schema by @looplj in https://github.com/looplj/axonhub/pull/1834
- feat: toggle IP ban icon by @nb5p in https://github.com/looplj/axonhub/pull/1848
- feat: add GLM-5.2 to zai channel default models by @qinkangdeid in https://github.com/looplj/axonhub/pull/1842
- fix(oidc): avoid privacy-deny error on every sign-in by @qinkangdeid in https://github.com/looplj/axonhub/pull/1841
- fix: model circuit breakder, close #1821 by @looplj in https://github.com/looplj/axonhub/pull/1854
- fix(model_fetcher): avoid duplicate /v1 for anthropic-like endpoints by @KremeCN in https://github.com/looplj/axonhub/pull/1852
- feat: add array_remove override op by @cesaryuan in https://github.com/looplj/axonhub/pull/1851
- feat: reset codex quota right now, close #1849 by @looplj in https://github.com/looplj/axonhub/pull/1855
- fix(frontend): correct api key creator filter scope by @KremeCN in https://github.com/looplj/axonhub/pull/1856
- fix: allow third-party Codex channels to edit Base URL by @brynne8 in https://github.com/looplj/axonhub/pull/1835
- fix: enforce strict local channel RPM admission by @lihaojiee in https://github.com/looplj/axonhub/pull/1840
- fix: openai video schema, close #1845 by @looplj in https://github.com/looplj/axonhub/pull/1857
- fix(frontend): hide api key creator column without system user permission by @KremeCN in https://github.com/looplj/axonhub/pull/1859
- feat: test all api keys, close #1853 by @looplj in https://github.com/looplj/axonhub/pull/1858
- opt: init channel performances async, close #1861 by @looplj in https://github.com/looplj/axonhub/pull/1864
- fix(frontend/dashboard): align channel success rate card height with daily overview by @lihaojiee in https://github.com/looplj/axonhub/pull/1862
- fix: should drop the invalid signature for unknown openai/anthropic, … by @looplj in https://github.com/looplj/axonhub/pull/1866
- fix: api key token usage scope, close #1865 by @looplj in https://github.com/looplj/axonhub/pull/1867
- feat: highlight low cache hit rate, close #1837 by @looplj in https://github.com/looplj/axonhub/pull/1868
- fix: hide prefixed model, close #1732 by @looplj in https://github.com/looplj/axonhub/pull/1869
- fix: codex image generation by @looplj in https://github.com/looplj/axonhub/pull/1870
- chore(deps-dev): bump vite from 7.3.2 to 7.3.5 in /frontend by @dependabot[bot] in https://github.com/looplj/axonhub/pull/1874
- feat: add transformer for cerebras, close #1777 by @looplj in https://github.com/looplj/axonhub/pull/1876
- fix: image generation bypass, close #1802 by @looplj in https://github.com

What's Changed
- chore: 调整图像输入 size 大小限制 by @sxueck in https://github.com/looplj/axonhub/pull/1769
- feat: add modalities field to /v1/models API response by @MoshiCoCo in https://github.com/looplj/axonhub/pull/1772
- feat: add IP blocklist settings by @hen7777777 in https://github.com/looplj/axonhub/pull/1771
- fix(frontend): guard execution channel query by permission by @MoeMion in https://github.com/looplj/axonhub/pull/1770
- opt: backup/restore usages logs seprated, close #1712 by @looplj in https://github.com/looplj/axonhub/pull/1766
- chore: upgrade claude code headers by @looplj in https://github.com/looplj/axonhub/pull/1774
- fix: allow users to update language preference by @evanhon0110 in https://github.com/looplj/axonhub/pull/1782
- fix: cache did not refresh sometime by @looplj in https://github.com/looplj/axonhub/pull/1783
- fix(datastorage): remove unbounded in-memory cache of request/response bodies (OOM) by @suixinio in https://github.com/looplj/axonhub/pull/1781
- fix(llm): claude code quota check header by @looplj in https://github.com/looplj/axonhub/pull/1785
- feat(openapi): add apiKeyQuotaUsages query for service-account keys by @suixinio in https://github.com/looplj/axonhub/pull/1779
- fix(openai): aggregate stream choices by index to avoid nil panic by @SAY-5 in https://github.com/looplj/axonhub/pull/1775
- chore: upgrade quic by @looplj in https://github.com/looplj/axonhub/pull/1786
- feat: add OpenAI audio APIs (TTS / STT) with SSE streaming by @EmccK in https://github.com/looplj/axonhub/pull/1784
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1795
- fix(audio): 修复 TTS 流式 Accept、外部存储与持久化内存占用 by @EmccK in https://github.com/looplj/axonhub/pull/1792
- feat(channels): adding Anthropic messages API format support to the OpenCodeGo channel (#1751) by @MoshiCoCo in https://github.com/looplj/axonhub/pull/1791
- feat: support redis multi mode by @wangYX657211334 in https://github.com/looplj/axonhub/pull/1788
- refactor(neuralwatt): replace current_period_start/end with kwh_reset_date by @djdembeck in https://github.com/looplj/axonhub/pull/1794
- fix: preserve Responses function call namespace by @DevilSi0722 in https://github.com/looplj/axonhub/pull/1797
- feat: 添加响应超时重试配置 by @xuyufengfei in https://github.com/looplj/axonhub/pull/1804
- feat: add evolink provider by @EvoLinkAI in https://github.com/looplj/axonhub/pull/1773
- feat(quota): add Apertis quota monitoring and refactor openai_responses handling by @djdembeck in https://github.com/looplj/axonhub/pull/1793
- feat: channel retryable statu codes, close #1803 by @looplj in https://github.com/looplj/axonhub/pull/1807
- feat: route with media condition, close #1729 by @looplj in https://github.com/looplj/axonhub/pull/1809
New Contributors
- @MoeMion made their first contribution in https://github.com/looplj/axonhub/pull/1770
- @evanhon0110 made their first contribution in https://github.com/looplj/axonhub/pull/1782
- @SAY-5 made their first contribution in https://github.com/looplj/axonhub/pull/1775
- @EmccK made their first contribution in https://github.com/looplj/axonhub/pull/1784
- @DevilSi0722 made their first contribution in https://github.com/looplj/axonhub/pull/1797
- @xuyufengfei made their first contribution in https://github.com/looplj/axonhub/pull/1804
- @EvoLinkAI made their first contribution in https://github.com/looplj/axonhub/pull/1773
Full Changelog: https://github.com/looplj/axonhub/compare/v1.0.0-beta2...v1.0.0-beta3

What's Changed
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1706
- fix: scroll for the channel rate limit dialog, close #1701 by @looplj in https://github.com/looplj/axonhub/pull/1708
- feat: render structured request log content, close #1698 by @looplj in https://github.com/looplj/axonhub/pull/1711
- fix: 修复了多图输入情况下转发失效的问题 by @sxueck in https://github.com/looplj/axonhub/pull/1714
- feat: add advanced model route conditions, close #1670 by @looplj in https://github.com/looplj/axonhub/pull/1717
- 修复 Responses 非透传模式下工具原语保真 by @deqiying in https://github.com/looplj/axonhub/pull/1703
- feat: support non-stream fallback to require-stream channels by @ttttmr in https://github.com/looplj/axonhub/pull/1661
- fix(llm): accept structured response item arguments by @llc1123 in https://github.com/looplj/axonhub/pull/1728
- fix(channels): show real i18n label for single-variant types like opencode_go by @qinkangdeid in https://github.com/looplj/axonhub/pull/1734
- chore: add claude-opus-4-7 to claudecode DefaultModels by @qinkangdeid in https://github.com/looplj/axonhub/pull/1733
- feat(llm): support responses websocket sessions by @llc1123 in https://github.com/looplj/axonhub/pull/1730
- Preserve request log filters on detail navigation by @zhaozhaozz in https://github.com/looplj/axonhub/pull/1723
- chore: 添加了登陆后页面标题自定义配置项 by @sxueck in https://github.com/looplj/axonhub/pull/1720
- fix: model settings dialog, close #1716 by @looplj in https://github.com/looplj/axonhub/pull/1718
- Support case-insensitive model filter patterns by @zhaozhaozz in https://github.com/looplj/axonhub/pull/1725
- chore: add claude-opus-4-8 to claudecode DefaultModels by @qinkangdeid in https://github.com/looplj/axonhub/pull/1739
- feat: opencode trace, close #1740 by @looplj in https://github.com/looplj/axonhub/pull/1744
- feat: reactivate api key, close #1735 by @looplj in https://github.com/looplj/axonhub/pull/1745
- fix: model association page throws Invalid input for channels without endpoints by @djdembeck in https://github.com/looplj/axonhub/pull/1746
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1754
- fix(anthropic): handle server-side tool blocks without panicking by @xdqi in https://github.com/looplj/axonhub/pull/1750
- opt: request detail sheet annimation, close #1749 by @looplj in https://github.com/looplj/axonhub/pull/1757
- fix: should set default model type if type not present, close #1752 by @looplj in https://github.com/looplj/axonhub/pull/1760
- feat: support disabling db auto migration by @wangYX657211334 in https://github.com/looplj/axonhub/pull/1727
- fix(frontend): avoid leaking "{{resource}}" placeholder in NOT_FOUND errors by @qinkangdeid in https://github.com/looplj/axonhub/pull/1762
- fix: auto reasoning effort for qwen max, close #1751 by @looplj in https://github.com/looplj/axonhub/pull/1765
New Contributors
- @sxueck made their first contribution in https://github.com/looplj/axonhub/pull/1714
- @xdqi made their first contribution in https://github.com/looplj/axonhub/pull/1750
- @wangYX657211334 made their first contribution in https://github.com/looplj/axonhub/pull/1727
Full Changelog: https://github.com/looplj/axonhub/compare/v1.0.0-beta1...v1.0.0-beta2

What's Changed
- fix(openai): tolerate object-shaped usage cost by @looplj in https://github.com/looplj/axonhub/pull/2373
- Pass the caller's context through the v0.3.0 and v0.4.0 migrations by @omlahore in https://github.com/looplj/axonhub/pull/2348
- fix(openai): preserve request contracts across transports by @techblack in https://github.com/looplj/axonhub/pull/2374
- fix(openai): prevent zero-status errors from returning HTTP 200 by @Solaris-star in https://github.com/looplj/axonhub/pull/2372
- fix(images): accept application/json in passthrough mode for /images/edits (#2336) by @remote-controlled-man in https://github.com/looplj/axonhub/pull/2370
- fix(anthropic): preserve client cache_control breakpoints to keep prompt cache hits by @Azhc in https://github.com/looplj/axonhub/pull/2342
- fix(trace): stop persisting traces for embedding requests by @trim21 in https://github.com/looplj/axonhub/pull/2378
- fix(frontend): silence transient channel polling errors by @llc1123 in https://github.com/looplj/axonhub/pull/2382
- fix(frontend): handle empty channel error messages by @llc1123 in https://github.com/looplj/axonhub/pull/2375
- fix(api): report server stream deadlines to SSE clients by @llc1123 in https://github.com/looplj/axonhub/pull/2362
- feat: add ZenMux channels and quota-aware UI by @llc1123 in https://github.com/looplj/axonhub/pull/2381
- fix: window deploy scripts, close #2376 by @looplj in https://github.com/looplj/axonhub/pull/2386
- fix(openai): complete cross-format request conversion by @techblack in https://github.com/looplj/axonhub/pull/2377
- fix(frontend): localize model developer selector by @llc1123 in https://github.com/looplj/axonhub/pull/2385
- docs: move deployment to axonhub-skills by @looplj in https://github.com/looplj/axonhub/pull/2388
- fix: include hy4 models in Tencent catalog by @imzrme in https://github.com/looplj/axonhub/pull/2387
- fix(openai): merge multiple system messages for strict upstreams by @itachaaa in https://github.com/looplj/axonhub/pull/2384
- feat: archive personal api keys when delete user, close #2272 by @looplj in https://github.com/looplj/axonhub/pull/2391
- feat: fire webhook when api key disabled, close #2380 by @looplj in https://github.com/looplj/axonhub/pull/2393
- feat: remove cc billing system message for non official channel, clos… by @looplj in https://github.com/looplj/axonhub/pull/2394
- feat: add Command Code channel support by @kaiwishc in https://github.com/looplj/axonhub/pull/2392
- fix: propagate OpenCode sessions to custom endpoints by @imzrme in https://github.com/looplj/axonhub/pull/2389
- feat: show OAuth channel quotas by @techblack in https://github.com/looplj/axonhub/pull/2366
- fix(zai): support GLM-5.2+ reasoning_effort and GLM-5.3 always-on thinking by @MrLiuGangQiang in https://github.com/looplj/axonhub/pull/2314
- feat(codex): support gpt-6-astra by @MrLiuGangQiang in https://github.com/looplj/axonhub/pull/2399
- fix: render image-generation traces without GraphQL errors by @mvanhorn in https://github.com/looplj/axonhub/pull/2297
- ci: enable manual unstable Docker image publishing by @llc1123 in https://github.com/looplj/axonhub/pull/2403
- chore(deps): bump google.golang.org/grpc from 1.82.1 to 1.83.1 in /integration_test/gemini by @dependabot[bot] in https://github.com/looplj/axonhub/pull/2404
- fix(requests): keep API key filter visible without results by @MrLiuGangQiang in https://github.com/looplj/axonhub/pull/2405
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/2412
- feat: add ZenMux native video generation by @llc1123 in https://github.com/looplj/axonhub/pull/2410
- fix(frontend): preserve channel form focus and onboarding dismissal by @llc1123 in https://github.com/looplj/axonhub/pull/2411
New Contributors
- @omlahore made their first contribution in https://github.com/looplj/axonhub/pull/2348
- @Solaris-star made their first contribution in https://github.com/looplj/axonhub/pull

What's Changed
- feat: 请求日志记录新增'推理强度'字段 by @deqiying in https://github.com/looplj/axonhub/pull/1506
- feat: per-channel concurrency queue + admission control by @suixinio in https://github.com/looplj/axonhub/pull/1503
- feat: auto reasoning effort with suffix by @looplj in https://github.com/looplj/axonhub/pull/1515
- feat: add optional read-write separation for database by @lazhenyi in https://github.com/looplj/axonhub/pull/1516
- feat(quota): add wafer, synthetic, and neuralwatt quota checkers by @djdembeck in https://github.com/looplj/axonhub/pull/1509
- fix(dashboard): align token stats by API key/channel/model with totals by @raikyou in https://github.com/looplj/axonhub/pull/1524
- feat: add Anthropic endpoints for Volcengine, AiHubMix, and Xiaomi MiMo by @banlanzs in https://github.com/looplj/axonhub/pull/1521
- display token usage in channel cards via tokenStatsByChannel API by @banlanzs in https://github.com/looplj/axonhub/pull/1528
- feat(quota): add provider quota enforcement with per-limit status tracking by @djdembeck in https://github.com/looplj/axonhub/pull/1518
- feat: customized channel endpoints by @looplj in https://github.com/looplj/axonhub/pull/1529
- fix(pipeline): prevent channels from getting stuck in processing status by @djdembeck in https://github.com/looplj/axonhub/pull/1530
- chore: reformat llm package by @looplj in https://github.com/looplj/axonhub/pull/1531
- fix(models): clicking Archive on a disabled model had no effect by @djdembeck in https://github.com/looplj/axonhub/pull/1535
- fix(ui): show used/total in NanoGPT quota display instead of used/remaining by @djdembeck in https://github.com/looplj/axonhub/pull/1543
- fix(models): resolve vision capability from modalities when explicit field is absent by @djdembeck in https://github.com/looplj/axonhub/pull/1542
- opt: disable spell check for API key textarea by @BLumia in https://github.com/looplj/axonhub/pull/1544
- opt: custom endpoint api format by @looplj in https://github.com/looplj/axonhub/pull/1545
- feat: align endpoint i18n and prompt protection preview by @looplj in https://github.com/looplj/axonhub/pull/1546
- feat(oidc): Initial OpenID Connect Support by @LazuliKao in https://github.com/looplj/axonhub/pull/1091
- scheduler: centralized cron job management by @looplj in https://github.com/looplj/axonhub/pull/1550
- chore: update issue template by @looplj in https://github.com/looplj/axonhub/pull/1551
- fix(llm): missing anthropic thinking disabled transform by @looplj in https://github.com/looplj/axonhub/pull/1552
- feat: openai/deepseek completion api, close #1458 by @looplj in https://github.com/looplj/axonhub/pull/1554
- chore: change prompt content column, close #1553 by @looplj in https://github.com/looplj/axonhub/pull/1580
- fix: max model association priority, close #1507 by @looplj in https://github.com/looplj/axonhub/pull/1581
- opt: responses stream failed handle & request executions index, close #1473 by @looplj in https://github.com/looplj/axonhub/pull/1585
- opt: openai outbound reasoning field config, close #1584 by @looplj in https://github.com/looplj/axonhub/pull/1586
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1588
- feat: global pass-through setting by @raikyou in https://github.com/looplj/axonhub/pull/1591
- feat: add lowercaseModelName channel setting for cross-channel load b… by @nb5p in https://github.com/looplj/axonhub/pull/1589
- feat: expose inbound request headers to override templates by @ttttmr in https://github.com/looplj/axonhub/pull/1520
- feat(apikeys): add API key profile templates by @djdembeck in https://github.com/looplj/axonhub/pull/1590
- feat: prefer outbound format equals to inbound format by @looplj in https://github.com/looplj/axonhub/pull/1592
- 修改前端，"测试场"新增测试模型网关模型 by @zhai23 in https://github.com/looplj/axonhub/pull/1597
- fix(gc): drop nil bind args so VACUUM actually runs on Postgres by @suixinio in https://github.com/looplj/axonh

Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.42...v0.9.43

Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.40...v0.9.42

Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.39...v0.9.40

What's Changed
- fix(list_models): respect API key profile filtering in full model data mode by @djdembeck in https://github.com/looplj/axonhub/pull/1480
- feat: success rate page for all channels by @banlanzs in https://github.com/looplj/axonhub/pull/1483
- feat: array override ops (array_append, array_prepend, array_insert) by @paopaoandlingyia in https://github.com/looplj/axonhub/pull/1481
- fix: deepseek reasoning content content aggregator by @looplj in https://github.com/looplj/axonhub/pull/1485
- chore(deps): bump go.opentelemetry.io/otel from 1.39.0 to 1.41.0 in /integration_test/gemini by @dependabot[bot] in https://github.com/looplj/axonhub/pull/1479
- feat(anthropic): perserve top level cache control, close #1484 by @looplj in https://github.com/looplj/axonhub/pull/1486
- fix: should not mark as completed if upstream error, close #1444 by @looplj in https://github.com/looplj/axonhub/pull/1488
- fix: force keep reasoning content for deepseek by @looplj in https://github.com/looplj/axonhub/pull/1491
- feat: add NanoGPT API quota monitoring support by @djdembeck in https://github.com/looplj/axonhub/pull/1490
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1494
- fix: should pass through response/stream, close #1442 by @looplj in https://github.com/looplj/axonhub/pull/1495
- feat: add stream condition to the model association by @looplj in https://github.com/looplj/axonhub/pull/1498
- opt: enable vacuum by default, close #1492 by @looplj in https://github.com/looplj/axonhub/pull/1500
- fix(xiaomi): map models provided mimo token plan as xiaomi models by @trim21 in https://github.com/looplj/axonhub/pull/1501
- feat: api key prefix config, close #1475 by @looplj in https://github.com/looplj/axonhub/pull/1502
- feat: deepseek anthropic adaptive thinking by @looplj in https://github.com/looplj/axonhub/pull/1505
New Contributors
- @paopaoandlingyia made their first contribution in https://github.com/looplj/axonhub/pull/1481
- @trim21 made their first contribution in https://github.com/looplj/axonhub/pull/1501
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.37...v0.9.38

Major Changes
1. Added new the db connection config
重大变更
1. 增加了数据库连接池设置
What's Changed
- feat(channels): add bailian_anthropic and moonshot_coding channel types with UI checkbox support by @banlanzs in https://github.com/looplj/axonhub/pull/1455
- fix: model alias from/to align, close #1454 by @looplj in https://github.com/looplj/axonhub/pull/1460
- opt: add database connection limit, close #1459 by @looplj in https://github.com/looplj/axonhub/pull/1462
- opt: block merge for hop-by-hop headers, close #1452 by @looplj in https://github.com/looplj/axonhub/pull/1463
- feat: import codex auth json, close #1425 by @looplj in https://github.com/looplj/axonhub/pull/1465
- fix: anthropic thinking block schema by @looplj in https://github.com/looplj/axonhub/pull/1466
- fix(ui): add missing field in secondary window by @LazuliKao in https://github.com/looplj/axonhub/pull/1457
- fix: fix external_id length by @LazuliKao in https://github.com/looplj/axonhub/pull/1472
- feat: github quota & refactor quota types by @LazuliKao in https://github.com/looplj/axonhub/pull/1470
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1474
- fix: should ensure anthropic thinking signature exit by @looplj in https://github.com/looplj/axonhub/pull/1477
- feat: bulk clear channel override oprations and allow replace when bu… by @looplj in https://github.com/looplj/axonhub/pull/1478
New Contributors
- @banlanzs made their first contribution in https://github.com/looplj/axonhub/pull/1455
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.36...v0.9.37

What's Changed
- fix(ui): show secondary rate limit reset time in quota badges by @myml in https://github.com/looplj/axonhub/pull/1437
- feat(channels): add ollama channel support by @myml in https://github.com/looplj/axonhub/pull/1439
- fix: channel probe filter & token stats by @looplj in https://github.com/looplj/axonhub/pull/1446
- fix: rerank pass through body model by @looplj in https://github.com/looplj/axonhub/pull/1447
- fix: empty detect for non chat request type, close #1445 by @looplj in https://github.com/looplj/axonhub/pull/1448
- feat: add Meta developer and Llama models support by @djdembeck in https://github.com/looplj/axonhub/pull/1449
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.35...v0.9.36

What's Changed
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1416
- Revert "fix(channels): bulk weight adjustment not populating on drag … by @looplj in https://github.com/looplj/axonhub/pull/1420
- doc: fix broken doc link by @looplj in https://github.com/looplj/axonhub/pull/1422
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1423
- docs: rename model mapping to model alias by @looplj in https://github.com/looplj/axonhub/pull/1433
- fix: force codex channel to use array inputs, close #1430 by @looplj in https://github.com/looplj/axonhub/pull/1436
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.34...v0.9.35

What's Changed
- opt: change save data to not return value by @looplj in https://github.com/looplj/axonhub/pull/1405
- opt: generate stable device id and account uuid for claude code by @looplj in https://github.com/looplj/axonhub/pull/1406
- fix(channels): bulk weight adjustment not populating on drag reorder by @djdembeck in https://github.com/looplj/axonhub/pull/1401
- chore(i18n): 补充 StepFun 开发者名称翻译 by @llc1123 in https://github.com/looplj/axonhub/pull/1407
- fix(anthropic): accept xhigh effort level for Claude Opus 4.7 by @qinkangdeid in https://github.com/looplj/axonhub/pull/1412
- feat: pass through request body, close #1383 by @looplj in https://github.com/looplj/axonhub/pull/1411
- fix: channel health probe sql, close #1408 by @looplj in https://github.com/looplj/axonhub/pull/1414
- doc: add doc index, close #1393 by @looplj in https://github.com/looplj/axonhub/pull/1415
New Contributors
- @qinkangdeid made their first contribution in https://github.com/looplj/axonhub/pull/1412
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.33...v0.9.34

What's Changed
- feat: gemini/doubao embedding, close #1093 by @looplj in https://github.com/looplj/axonhub/pull/1388
- opt: try to fix channel/model cache not refreshed by @looplj in https://github.com/looplj/axonhub/pull/1390
- opt: align codex trace header with latest codex cli by @yoke233 in https://github.com/looplj/axonhub/pull/1387
- feat: realtime chunks preview and response preview by @LazuliKao in https://github.com/looplj/axonhub/pull/1342
- feat: add cache diagnostics tool by @looplj in https://github.com/looplj/axonhub/pull/1394
- fix: unwrap DebugDriver layer in Driver() to fix perf analysis in debug model by @MoshiCoCo in https://github.com/looplj/axonhub/pull/1397
- refactor: use the live streaming middle for live preview by @looplj in https://github.com/looplj/axonhub/pull/1395
- feat: add empty strem response detection, close #292 by @looplj in https://github.com/looplj/axonhub/pull/1399
- opt: mobile ui for pages/dialogs, close #946 by @looplj in https://github.com/looplj/axonhub/pull/1400
New Contributors
- @yoke233 made their first contribution in https://github.com/looplj/axonhub/pull/1387
- @MoshiCoCo made their first contribution in https://github.com/looplj/axonhub/pull/1397
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.32...v0.9.33

What's Changed
- feat: channel rpm settings, close #746 by @looplj in https://github.com/looplj/axonhub/pull/1320
- feat: channel concurrency settings, close #1130 by @looplj in https://github.com/looplj/axonhub/pull/1322
- chore(deps): upgrade aws eventstream by @looplj in https://github.com/looplj/axonhub/pull/1332
- feat: add kwaipilot developer catalog support by @llc1123 in https://github.com/looplj/axonhub/pull/1328
- feat: repsect channel response with retry after, close #858 by @looplj in https://github.com/looplj/axonhub/pull/1333
- chore(i18n): 补充模型开发者本地化标签 by @llc1123 in https://github.com/looplj/axonhub/pull/1334
- feat: add latency aware strategy, close #962 by @looplj in https://github.com/looplj/axonhub/pull/1339
- feat: add image generation support to NanoGPT channel by @djdembeck in https://github.com/looplj/axonhub/pull/1345
- fix(codex): 修复下游无 system message 时 Codex 渠道返回 400 的问题 by @llc1123 in https://github.com/looplj/axonhub/pull/1346
- feat: Add configurable default response mode for /v1/models by @raikyou in https://github.com/looplj/axonhub/pull/1347
- chore(i18n): fix zh-CN name placeholders by @qaqland in https://github.com/looplj/axonhub/pull/1349
- opt: minor theme optimization by @looplj in https://github.com/looplj/axonhub/pull/1354
- feat: batch test channesl api keys, close #1318 by @looplj in https://github.com/looplj/axonhub/pull/1356
- fix: auto reg anchor, close #1355 by @looplj in https://github.com/looplj/axonhub/pull/1359
- fix: do not unescape json for request/reponse body, close #1357 by @looplj in https://github.com/looplj/axonhub/pull/1360
- fix: curl generator for gemini vertex, close #1329 by @looplj in https://github.com/looplj/axonhub/pull/1363
- feat: add webhook notification when channel disabled, close #1315 by @looplj in https://github.com/looplj/axonhub/pull/1365
- fix: quota check infifite loop by @looplj in https://github.com/looplj/axonhub/pull/1366
- feat: add proxy to webhook target by @looplj in https://github.com/looplj/axonhub/pull/1367
- fix(anthropic): derive prompt cache key from request context by @zccrs in https://github.com/looplj/axonhub/pull/1350
- feat: allow view test history by @looplj in https://github.com/looplj/axonhub/pull/1371
- fix: should filter out custom tool for non response channel, close #1370 by @looplj in https://github.com/looplj/axonhub/pull/1372
- fix(copilot): infer X-Initiator from message role for quota tracking by @djdembeck in https://github.com/looplj/axonhub/pull/1375
- feat: add nanogpt_responses channel type by @djdembeck in https://github.com/looplj/axonhub/pull/1376
- feat: add condition to the model association, close #535 by @looplj in https://github.com/looplj/axonhub/pull/1379
- feat: project/apikey profile allow exclude channel tags, close #1331 by @looplj in https://github.com/looplj/axonhub/pull/1382
New Contributors
- @qaqland made their first contribution in https://github.com/looplj/axonhub/pull/1349
- @zccrs made their first contribution in https://github.com/looplj/axonhub/pull/1350
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.31...v0.9.32

What's Changed
- fix: allow nil project profiles, close #1316 by @looplj in https://github.com/looplj/axonhub/pull/1317
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.30...v0.9.31

What's Changed
- docs: add Japanese README by @eltociear in https://github.com/looplj/axonhub/pull/1299
- feat: add project profiles for channel filtering close #1233 by @looplj in https://github.com/looplj/axonhub/pull/1303
- feat: search api key by key, close #1301 by @looplj in https://github.com/looplj/axonhub/pull/1305
- chore: clenup unnesessary project id for oauth channels by @looplj in https://github.com/looplj/axonhub/pull/1306
- opt: add timeout for fx startup, close #1304 by @looplj in https://github.com/looplj/axonhub/pull/1307
- opt: quota bdage float window, close #1300 by @looplj in https://github.com/looplj/axonhub/pull/1308
- feat: add NVIDIA as model provider by @djdembeck in https://github.com/looplj/axonhub/pull/1309
- chore(deps-dev): bump vite from 7.3.1 to 7.3.2 in /frontend by @dependabot[bot] in https://github.com/looplj/axonhub/pull/1310
- opt(responses): fix function parameters missing properties, close #1032 by @looplj in https://github.com/looplj/axonhub/pull/1311
- fix: project cache cause auth faield, close #1298 by @looplj in https://github.com/looplj/axonhub/pull/1312
- chore: cleanup agent rules by @looplj in https://github.com/looplj/axonhub/pull/1314
New Contributors
- @eltociear made their first contribution in https://github.com/looplj/axonhub/pull/1299
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.29...v0.9.30

What's Changed
- fix: missing translation and script by @looplj in https://github.com/looplj/axonhub/pull/1293
- fix: check translation key script and add missing i18n key, close #1295 by @looplj in https://github.com/looplj/axonhub/pull/1296
- doc: add sqlite wal docs by @looplj in https://github.com/looplj/axonhub/pull/1297
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.28...v0.9.29

What's Changed
- fix(moonshot): normalize negative input tokens by @djdembeck in https://github.com/looplj/axonhub/pull/1276
- ci: add build workflow by @djdembeck in https://github.com/looplj/axonhub/pull/1275
- doc: anouncement for claude code by @looplj in https://github.com/looplj/axonhub/pull/1280
- feat: allow delete user by @looplj in https://github.com/looplj/axonhub/pull/1284
- feat: allow delete project by @looplj in https://github.com/looplj/axonhub/pull/1285
- opt: threads/trace ui by @looplj in https://github.com/looplj/axonhub/pull/1289
- refactor: add staleness-based cleanup for stuck processing records by @djdembeck in https://github.com/looplj/axonhub/pull/1286
- feat: add flow/tree view for trace by @looplj in https://github.com/looplj/axonhub/pull/1290
- feat: filter requests by source by @looplj in https://github.com/looplj/axonhub/pull/1291
- refactor: cleanup i18n keys & mutation error message by @looplj in https://github.com/looplj/axonhub/pull/1292
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.27...v0.9.28

What's Changed
- fix: cmd/schema build failed by @looplj in https://github.com/looplj/axonhub/pull/1273
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.26...v0.9.27

What's Changed
- feat: add user agent pass-through settings by @djdembeck in https://github.com/looplj/axonhub/pull/1231
- feat(system): add prefer language during initialization by @myml in https://github.com/looplj/axonhub/pull/1236
- opt: do not route to static assets for API path by @looplj in https://github.com/looplj/axonhub/pull/1237
- chore: cleanup xxx ptr to use lo by @looplj in https://github.com/looplj/axonhub/pull/1238
- feat: api key profile channel tags compposite, close #1220 by @looplj in https://github.com/looplj/axonhub/pull/1240
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.24...v0.9.25

What's Changed
- refactor: 登录页背景动画循环并补充调试入口 by @llc1123 in https://github.com/looplj/axonhub/pull/1221
- fix: passthrough X-Initiator header in Copilot channel for billing control by @JasonWenTheFox in https://github.com/looplj/axonhub/pull/1222
- feat: bulk test channels, close #1219 by @looplj in https://github.com/looplj/axonhub/pull/1228
- fix(nanogpt): implement XML tool call parsing for NanoGPT transformer by @djdembeck in https://github.com/looplj/axonhub/pull/1211
- feat: add prompt protect rules preview, close #1215 by @looplj in https://github.com/looplj/axonhub/pull/1229
New Contributors
- @JasonWenTheFox made their first contribution in https://github.com/looplj/axonhub/pull/1222
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.23...v0.9.24

What's Changed
- chore(codex): upgrade codex ua version by @looplj in https://github.com/looplj/axonhub/pull/1151
- doc: add request processing by @looplj in https://github.com/looplj/axonhub/pull/1155
- fix: auto disble text, close #1156 by @looplj in https://github.com/looplj/axonhub/pull/1159
- chore: 为 Codex 计划补充 gpt-5.3-codex-spark 和 gpt-5.4-mini 模型 by @llc1123 in https://github.com/looplj/axonhub/pull/1162
- refactor: 去除 Codex CLI 伪装并简化 Codex 兼容逻辑 by @llc1123 in https://github.com/looplj/axonhub/pull/1166
- feat: add Fireworks AI channel support by @djdembeck in https://github.com/looplj/axonhub/pull/1200
- feat(llm): response compact API by @looplj in https://github.com/looplj/axonhub/pull/1161
- fix: restore channel losing sync models pattern, close #1203 by @looplj in https://github.com/looplj/axonhub/pull/1205
- fix(channels): prevent ChannelsTypeTabs from wrapping on small screens by @nb5p in https://github.com/looplj/axonhub/pull/1206
- fix: doubao anthropic fetch models by @looplj in https://github.com/looplj/axonhub/pull/1207
- opt: allow invalid api key if allow no auth, close #1201 by @looplj in https://github.com/looplj/axonhub/pull/1208
- fix: duplicate override path when apply template by @looplj in https://github.com/looplj/axonhub/pull/1209
- feat: prompt inject condition for api key, close #1202 by @looplj in https://github.com/looplj/axonhub/pull/1210
New Contributors
- @llc1123 made their first contribution in https://github.com/looplj/axonhub/pull/1162
- @nb5p made their first contribution in https://github.com/looplj/axonhub/pull/1206
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.22...v0.9.23

What's Changed
- feat: standardize chart legend lists by @djdembeck in https://github.com/looplj/axonhub/pull/1141
- build: add freebsd support for releases by @cikichen in https://github.com/looplj/axonhub/pull/1146
- feat: add cli + skill, close #1085 by @looplj in https://github.com/looplj/axonhub/pull/1149
- fix: anthropic adaptive thinking effort did not pass to other outbound by @looplj in https://github.com/looplj/axonhub/pull/1150
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.21...v0.9.22

What's Changed
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1122
- fix(gemini): remove metadata and disable inbound query merge by @rainive in https://github.com/looplj/axonhub/pull/1124
- feat(dashboard): add time period filter to all charts by @djdembeck in https://github.com/looplj/axonhub/pull/1129
- fix(anthropic): empty thinking cause content block not stop by @looplj in https://github.com/looplj/axonhub/pull/1135
- fix(ci): sync model workflow should based on the unstable branch, close #1125 by @looplj in https://github.com/looplj/axonhub/pull/1136
- chore(deps): bump google.golang.org/grpc from 1.77.0 to 1.79.3 in /integration_test/gemini by @dependabot[bot] in https://github.com/looplj/axonhub/pull/1138
- chore: update issue template by @looplj in https://github.com/looplj/axonhub/pull/1140
New Contributors
- @rainive made their first contribution in https://github.com/looplj/axonhub/pull/1124
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.20...v0.9.21

What's Changed
- feat: allow no auth for api, close #1104 by @looplj in https://github.com/looplj/axonhub/pull/1107
- feat: prompt protect, close #850 wip by @looplj in https://github.com/looplj/axonhub/pull/1109
- fix(anthropic): omitted thinking block, close #1105 by @looplj in https://github.com/looplj/axonhub/pull/1113
- fix(anthropic): panic for some cases, close #1112 by @looplj in https://github.com/looplj/axonhub/pull/1114
- fix: no auth and prompt protect, misc by @looplj in https://github.com/looplj/axonhub/pull/1116
- chore(deps): bump google.golang.org/grpc from 1.77.0 to 1.79.3 in /llm by @dependabot[bot] in https://github.com/looplj/axonhub/pull/1115
- feat: video input, close #892 by @looplj in https://github.com/looplj/axonhub/pull/1117
- feat: audio input/output by @looplj in https://github.com/looplj/axonhub/pull/1118
- opt: do not block for load font, close #1089 by @looplj in https://github.com/looplj/axonhub/pull/1119
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.19...v0.9.20

What's Changed
- feat(apikeys): add token usage statistics with read permission support by @hen7777777 in https://github.com/looplj/axonhub/pull/1087
- opt(llm): thinking signature encode with channel footprint, close #1035 by @looplj in https://github.com/looplj/axonhub/pull/1066
- chore(deps): bump google.golang.org/grpc from 1.77.0 to 1.79.3 by @dependabot[bot] in https://github.com/looplj/axonhub/pull/1096
- doc: add note to readme by @looplj in https://github.com/looplj/axonhub/pull/1099
- fix(llm): claude code user id compatible by @looplj in https://github.com/looplj/axonhub/pull/1102
- feat: request/cost/token by channel/model/apikey, close #1097 by @looplj in https://github.com/looplj/axonhub/pull/1106
New Contributors
- @hen7777777 made their first contribution in https://github.com/looplj/axonhub/pull/1087
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.18...v0.9.19

What's Changed
- feat(llm): add Reasoning field support for additional provider compatibility by @djdembeck in https://github.com/looplj/axonhub/pull/1077
- feat: add copy url button on api key dialog, close 1072 by @looplj in https://github.com/looplj/axonhub/pull/1079
- feat: save channel proxy configs, close #1078 by @looplj in https://github.com/looplj/axonhub/pull/1080
- feat: show request execution error status code, close #1049 by @looplj in https://github.com/looplj/axonhub/pull/1082
- feat: disable ssl verification, close #1024 by @looplj in https://github.com/looplj/axonhub/pull/1084
- feat: add Homebrew tap support for automated formula publishing by @cikichen in https://github.com/looplj/axonhub/pull/1086
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.17...v0.9.18

What's Changed
- fix: preserve JSON-like string values in header override by @GhostFlying in https://github.com/looplj/axonhub/pull/1052
- opt: image display in request/trace, close #1047 by @looplj in https://github.com/looplj/axonhub/pull/1056
- fix(dashboard): incorrect time conversion in MySQL by @imlonghao in https://github.com/looplj/axonhub/pull/1058
- fix(orchestrator): preserve completed responses on client disconnect by @Chengxiwei in https://github.com/looplj/axonhub/pull/1059
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/1063
- fix(github-copilot): GPT-5.4 should use Responses API by @imlonghao in https://github.com/looplj/axonhub/pull/1064
New Contributors
- @GhostFlying made their first contribution in https://github.com/looplj/axonhub/pull/1052
- @imlonghao made their first contribution in https://github.com/looplj/axonhub/pull/1058
- @Chengxiwei made their first contribution in https://github.com/looplj/axonhub/pull/1059
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.16...v0.9.17

What's Changed
- fix: gemini 转换中 lastPart 为 nil 时避免 panic by @MomentDerek in https://github.com/looplj/axonhub/pull/1051
- Revert "opt: reduce sqlite "database is locked" error (#1009)" by @looplj in https://github.com/looplj/axonhub/pull/1053
- Revert "opt: auto reorder channels, close #822 (#1005)" by @looplj in https://github.com/looplj/axonhub/pull/1055
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.15...v0.9.16

What's Changed
- fix: sqlite "database is locked" error by @myml in https://github.com/looplj/axonhub/pull/1009
- feat: refine GC cleanup logic to support manual full purge by @cikichen in https://github.com/looplj/axonhub/pull/1010
- refactor: rename GraphQL mutation to triggerGcCleanup (follow-up to #1010) by @cikichen in https://github.com/looplj/axonhub/pull/1017
- feat: unlock codex stream limit by @PA733 in https://github.com/looplj/axonhub/pull/1019
- fix(biz): resolve WebDAV 405 error on Synology NAS by normalizing paths by @cikichen in https://github.com/looplj/axonhub/pull/1014
- fix: curl with duplicated version, close #1018 by @looplj in https://github.com/looplj/axonhub/pull/1020
- chore: add sql default value for time, close #1013 by @looplj in https://github.com/looplj/axonhub/pull/1021
- opt: disalbed channel notice, close #1025 by @looplj in https://github.com/looplj/axonhub/pull/1026
- opt: should not panic if the metrics type is empty, close #990 by @looplj in https://github.com/looplj/axonhub/pull/1027
- fix: should pass through the stream error, close #948 by @looplj in https://github.com/looplj/axonhub/pull/1031
- opt: always return the configured models, close #919 by @looplj in https://github.com/looplj/axonhub/pull/1038
- doc: update model settings guideline, close #1036 by @looplj in https://github.com/looplj/axonhub/pull/1039
- opt: improve compatible with error code, close #1037 by @looplj in https://github.com/looplj/axonhub/pull/1040
- opt: chat process does not need any extra scopes, close #1033 by @looplj in https://github.com/looplj/axonhub/pull/1041
- opt: allow fetch models for third part codex/cc channel by @looplj in https://github.com/looplj/axonhub/pull/1042
- opt: longcat request compatibity, close #1032 by @looplj in https://github.com/looplj/axonhub/pull/1044
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.14...v0.9.15

What's Changed
- feat: allow type editing for third-party codex/claudecode channels and make channel name clickable by @raikyou in https://github.com/looplj/axonhub/pull/992
- fix(gemini): response json schema compatible, close #991 by @looplj in https://github.com/looplj/axonhub/pull/1000
- feat(system): add manual log cleanup trigger UI by @cikichen in https://github.com/looplj/axonhub/pull/1001
- opt: auto reorder channels, close #822 by @looplj in https://github.com/looplj/axonhub/pull/1005
- feat: add gpt-5.4 to codex, close #999 by @looplj in https://github.com/looplj/axonhub/pull/1006
- feat: filter requests by model id, close #994 by @looplj in https://github.com/looplj/axonhub/pull/1007
New Contributors
- @cikichen made their first contribution in https://github.com/looplj/axonhub/pull/1001
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.13...v0.9.14

What's Changed
- feat(dashboard): add all-time token statistics by @djdembeck in https://github.com/looplj/axonhub/pull/945
- feat(requests): add quick-view drawer and improve JSON display by @raikyou in https://github.com/looplj/axonhub/pull/978
- feat(channels,models): allow editing provider, API format, developer, model ID, and type by @raikyou in https://github.com/looplj/axonhub/pull/982
- style: remove width limit on model list by @myml in https://github.com/looplj/axonhub/pull/983
- feat: auto fetch the latest version of antigravity for UA by @PA733 in https://github.com/looplj/axonhub/pull/984
- opt: refresh qutoa with channel proxy setting, close #926 by @looplj in https://github.com/looplj/axonhub/pull/986
New Contributors
- @raikyou made their first contribution in https://github.com/looplj/axonhub/pull/978
- @myml made their first contribution in https://github.com/looplj/axonhub/pull/983
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.12...v0.9.13

What's Changed
- opt(openai): filter out response custom tool for openai outbound by @looplj in https://github.com/looplj/axonhub/pull/954
- opt: channel override template default value, close #958 by @looplj in https://github.com/looplj/axonhub/pull/963
- chore: add setup script, close #925 by @looplj in https://github.com/looplj/axonhub/pull/964
- fix: should exit if the subcommand exits by @looplj in https://github.com/looplj/axonhub/pull/965
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.11...v0.9.12

What's Changed
- fix: cap performance chart display to prevent outlier skew by @djdembeck in https://github.com/looplj/axonhub/pull/941
- fix(anthropic): should not validate output config, close #949 by @looplj in https://github.com/looplj/axonhub/pull/950
- opt: version checker skip other service by @looplj in https://github.com/looplj/axonhub/pull/951
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.10...v0.9.11

What's Changed
- feat(channel): add GitHub Copilot with OAuth device flow support by @djdembeck in https://github.com/looplj/axonhub/pull/900
- Fix AutoSyncModelPattern not being saved in channel updates by @caglareker in https://github.com/looplj/axonhub/pull/917
- fix: sync models now with pattern, close #920 by @looplj in https://github.com/looplj/axonhub/pull/921
- refactor: move copilot token exchanger to the copilot internal by @looplj in https://github.com/looplj/axonhub/pull/924
- 兼容 OpenAI 入站 extra_fields.extra_content 的 thought_signature 解析，并修复签名 base64 不正确导致的 genai 报错问题 by @MomentDerek in https://github.com/looplj/axonhub/pull/908
- fix(orchestrator): detect complete responses without terminal events for Codex + NanoGPT by @djdembeck in https://github.com/looplj/axonhub/pull/915
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/928
- opt: signature compatible when switch channel in the same session by @looplj in https://github.com/looplj/axonhub/pull/931
- feat(channel): add Gemini Vertex model fetching via PublicProviderConf by @djdembeck in https://github.com/looplj/axonhub/pull/932
- fix(frontend): correct tokens per second calculation for cache hits by @djdembeck in https://github.com/looplj/axonhub/pull/934
- opt(anthropic): optimize adaptive thinking compatible by @looplj in https://github.com/looplj/axonhub/pull/940
- fix: llm lincese and readme by @looplj in https://github.com/looplj/axonhub/pull/943
New Contributors
- @caglareker made their first contribution in https://github.com/looplj/axonhub/pull/917
- @MomentDerek made their first contribution in https://github.com/looplj/axonhub/pull/908
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.9...v0.9.10

What's Changed
- opt: add space after data: for sse event by @looplj in https://github.com/looplj/axonhub/pull/890
- opt: model list group by developer, close #840 by @looplj in https://github.com/looplj/axonhub/pull/891
- fix(llm): missing response format json schema, close #893 by @looplj in https://github.com/looplj/axonhub/pull/897
- feat: auto sync models with filter, close #896 by @looplj in https://github.com/looplj/axonhub/pull/898
- feat: openai/doubao video generation by @looplj in https://github.com/looplj/axonhub/pull/899
- feat: add image/video generation model by @looplj in https://github.com/looplj/axonhub/pull/903
- fix: should block cf headers, close #901 by @looplj in https://github.com/looplj/axonhub/pull/904
- opt: auto complete performance by @looplj in https://github.com/looplj/axonhub/pull/905
- refactor(llm): unify llm usage model by @looplj in https://github.com/looplj/axonhub/pull/906
- opt: playground channel/model selection by @looplj in https://github.com/looplj/axonhub/pull/913
- chore(deps): bump filippo.io/edwards25519 from 1.1.0 to 1.1.1 by @dependabot[bot] in https://github.com/looplj/axonhub/pull/868
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.8...v0.9.9

What's Changed
- chore: add restart and upgrade script by @looplj in https://github.com/looplj/axonhub/pull/869
- fix: should skip onbording for non-owner users by @looplj in https://github.com/looplj/axonhub/pull/870
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/877
- fix(gemini): resolve Vertex AI function call ID and grouping issues (#874) by @djdembeck in https://github.com/looplj/axonhub/pull/880
- feat(dashboard): add performance graphs for models and channels by @djdembeck in https://github.com/looplj/axonhub/pull/872
- feat: add tool_choice, adaptive thinking and output_config support for Anthropic transformer by @orzmoe in https://github.com/looplj/axonhub/pull/882
- fix(llm): skip add cache control if requests without cache control, c… by @looplj in https://github.com/looplj/axonhub/pull/884
- fix(dashboard): remove overflow-hidden to fix card clipping on hover by @djdembeck in https://github.com/looplj/axonhub/pull/885
- fix(requests): use defaultValue fallback for column toggle translations by @djdembeck in https://github.com/looplj/axonhub/pull/886
- fix(ui): ensure active tabs use white text on orange background in Claude dark theme by @djdembeck in https://github.com/looplj/axonhub/pull/887
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/888
- opt(llm): do not forward axonhub customzied header to upstream, close… by @looplj in https://github.com/looplj/axonhub/pull/889
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.7...v0.9.8

What's Changed
- fix: auto-refresh pagination limit by @djdembeck in https://github.com/looplj/axonhub/pull/845
- opt: remove wal for sqlite in install scripts by @looplj in https://github.com/looplj/axonhub/pull/847
- refactor: cleanup llm package deps by @looplj in https://github.com/looplj/axonhub/pull/848
- refactor: migrate log in llm to slog by @looplj in https://github.com/looplj/axonhub/pull/849
- refactor: make llm a go mod by @looplj in https://github.com/looplj/axonhub/pull/851
- fix: block/sensitive headers canonical form by @9bingyin in https://github.com/looplj/axonhub/pull/853
- feat: allow delete channel override template, close #852 by @looplj in https://github.com/looplj/axonhub/pull/855
- fix(antigravity): repair broken OAuth authentication - remove API key fallback by @djdembeck in https://github.com/looplj/axonhub/pull/859
- feat(requests): make all columns toggleable in View menu by @djdembeck in https://github.com/looplj/axonhub/pull/862
- feat(llm): support responses custom tool, close #860 by @looplj in https://github.com/looplj/axonhub/pull/865
- opt: strip cc billing header cch in inbound by @looplj in https://github.com/looplj/axonhub/pull/866
- opt: anthropic inbound compatible with gpt encryped content by @looplj in https://github.com/looplj/axonhub/pull/867
New Contributors
- @9bingyin made their first contribution in https://github.com/looplj/axonhub/pull/853
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.6...v0.9.7

What's Changed
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/839
- fix: handle embedding responses for stats display by @djdembeck in https://github.com/looplj/axonhub/pull/835
- feat(claudecode): add IsOfficial param to strip billing header for non-OAuth channels by @looplj in https://github.com/looplj/axonhub/pull/842
- perf(channels): optimize model list rendering with virtual scrolling by @looplj in https://github.com/looplj/axonhub/pull/843
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.5...v0.9.6

重大变更
1. 重构了权限控制，如果遇到问题，欢迎及时反馈
MAJOR Changes
2. Refactored scopes control，feel free to feedback
What's Changed
- feat(dashboard): Add Fastest Performers Cards with Token/s Visualization by @djdembeck in https://github.com/looplj/axonhub/pull/802
- fix: token statistics card responsive layout overflow on mobile by @djdembeck in https://github.com/looplj/axonhub/pull/803
- refactor: introduce authz to unify bypass by @looplj in https://github.com/looplj/axonhub/pull/805
- fix: improve mobile header accessibility and sidebar navigation by @djdembeck in https://github.com/looplj/axonhub/pull/806
- chore: fix e2e tests by @looplj in https://github.com/looplj/axonhub/pull/808
- fix: load data stroage and fastest query args by @looplj in https://github.com/looplj/axonhub/pull/809
- opt: dashboard collapsible section by @looplj in https://github.com/looplj/axonhub/pull/811
- feat: persist dashboard section collapsed states in localStorage by @djdembeck in https://github.com/looplj/axonhub/pull/813
- fix: correct development guide link and add NanoGPT provider to README by @djdembeck in https://github.com/looplj/axonhub/pull/814
- feat(requests): Add tokens/s toggle to latency column by @djdembeck in https://github.com/looplj/axonhub/pull/816
- feat: auto vacuum after gc by @looplj in https://github.com/looplj/axonhub/pull/823
- fix: openai stream with gemini reasoning content by @looplj in https://github.com/looplj/axonhub/pull/824
- fix: avoid error message when init redirect by @looplj in https://github.com/looplj/axonhub/pull/825
- feat: add order to the prompt, close #821 by @looplj in https://github.com/looplj/axonhub/pull/826
- feat(channel): preserve manual models during sync by @djdembeck in https://github.com/looplj/axonhub/pull/828
- opt: system menu and page title by @looplj in https://github.com/looplj/axonhub/pull/832
- opt: filter out native tools for anthropic compatible channel by @looplj in https://github.com/looplj/axonhub/pull/833
- fix: should not cancel request if the stream completed, close #827 by @looplj in https://github.com/looplj/axonhub/pull/834
- feat: add proxy when create channel by @looplj in https://github.com/looplj/axonhub/pull/838
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.4...v0.9.5

What's Changed
- feat: disable model association by @looplj in https://github.com/looplj/axonhub/pull/788
- feat: Add Helm Chart support for Kubernetes deployment by @lazhenyi in https://github.com/looplj/axonhub/pull/789
- fix: add the missing gemini request fields by @looplj in https://github.com/looplj/axonhub/pull/793
- feat: auto optimize anthropic cache control by @orzmoe in https://github.com/looplj/axonhub/pull/780
- feat: add Xiaomi MIMO channel by @lazhenyi in https://github.com/looplj/axonhub/pull/797
New Contributors
- @lazhenyi made their first contribution in https://github.com/looplj/axonhub/pull/789
- @orzmoe made their first contribution in https://github.com/looplj/axonhub/pull/780
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.3...v0.9.4

重大变更
1. 支持单渠道多 keys，可能会有些不稳定，遇到问题可以随时反馈
2. 重构了参数覆盖，尽量兼容了已有数据，如果有不兼容了，可以随时反馈
MAJOR CHANGES
1. Multiple api keys per channel, maybe not stable, welcome feedback.
2. Refactor request override, legacy data keep compatible, feel free to file issues if something not work as expected.
What's Changed
- chore: remove txn for update channel id when retry by @looplj in https://github.com/looplj/axonhub/pull/724
- refactor: cleanup channel performance by @looplj in https://github.com/looplj/axonhub/pull/726
- refactor: transformer use api key provider instead of static api key by @looplj in https://github.com/looplj/axonhub/pull/727
- release: release v0.9.2 by @looplj in https://github.com/looplj/axonhub/pull/730
- feat: multiple api keys for one channel by @looplj in https://github.com/looplj/axonhub/pull/731
- refactor: cleanup channel credentials by @looplj in https://github.com/looplj/axonhub/pull/732
- fix: dialog close button position by @looplj in https://github.com/looplj/axonhub/pull/733
- fix: force reasoning summary presence by @x22x22 in https://github.com/looplj/axonhub/pull/735
- fix(frontend): improve mobile responsive layouts by @duoduohe in https://github.com/looplj/axonhub/pull/737
- feat: trace sticky api key for multiple api keys channel by @looplj in https://github.com/looplj/axonhub/pull/740
- fix(anthropic): support image content blocks in tool_result by @duoduohe in https://github.com/looplj/axonhub/pull/742
- fix: remove codex metadata field by @looplj in https://github.com/looplj/axonhub/pull/745
- fix: daily request may missing data from today by @looplj in https://github.com/looplj/axonhub/pull/747
- fix: s3 path style storage, close #736 by @looplj in https://github.com/looplj/axonhub/pull/748
- opt: logging channel for the failed request, close #682 by @looplj in https://github.com/looplj/axonhub/pull/749
- opt: ignore deleted and no enabled api keys channel by @looplj in https://github.com/looplj/axonhub/pull/750
- opt: tester, model associate, tools by @looplj in https://github.com/looplj/axonhub/pull/751
- feat(api): enhance models endpoint by @djdembeck in https://github.com/looplj/axonhub/pull/743
- fix: preserve stream flag for TTFT persistence by @djdembeck in https://github.com/looplj/axonhub/pull/753
- fix(channel): use completion tokens for TPS calculation instead of total tokens by @djdembeck in https://github.com/looplj/axonhub/pull/755
- fix: responses encrypted content convertsion, and related tools by @looplj in https://github.com/looplj/axonhub/pull/758
- fix: channel test model not found, close #754 by @looplj in https://github.com/looplj/axonhub/pull/759
- doc: update the token chinese translation by @looplj in https://github.com/looplj/axonhub/pull/760
- feat: add NanoGPT provider support by @djdembeck in https://github.com/looplj/axonhub/pull/761
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/765
- fix: respect user timezone in daily overview widget by @djdembeck in https://github.com/looplj/axonhub/pull/766
- opt: avoid to read closed stream by @looplj in https://github.com/looplj/axonhub/pull/767
- fix: requests page filtering not working with auto refresh enabled by @djdembeck in https://github.com/looplj/axonhub/pull/768
- feat: support minute-level API key quota limiting; display historical quota usage by @TundraWork in https://github.com/looplj/axonhub/pull/770
- refactor: refactor channel override to structure, close #762 by @looplj in https://github.com/looplj/axonhub/pull/771
- refactor: channel override tempalte by @looplj in https://github.com/looplj/axonhub/pull/773
- opt: model fetch, auto summary and duplicate by @looplj in https://github.com/looplj/axonhub/pull/775
- feat: hide archived items in requests filters by default by @djdembeck in https://github.com/looplj/axonhub/pull/769
- opt: backup/restore for model association and api key profile by @looplj in https://github.com/looplj/axonhub/pull/776
- opt: multiple api

What's Changed
- fix: allow custom base URL for codex/claudecode third-party auth by @yuancjun in https://github.com/looplj/axonhub/pull/729
New Contributors
- @yuancjun made their first contribution in https://github.com/looplj/axonhub/pull/729
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.1...v0.9.2

What's Changed
- fix: failed to backup to gcs/s3, close #719 by @looplj in https://github.com/looplj/axonhub/pull/720
- opt: dialog compitable for vertical tabs, close #713 by @looplj in https://github.com/looplj/axonhub/pull/722
- feat: s3 storage path style, close #721 by @looplj in https://github.com/looplj/axonhub/pull/723
Full Changelog: https://github.com/looplj/axonhub/compare/v0.9.0...v0.9.1

BREAKING CHANGES
1. removed image generation via chat api, use the OpenAI image generation endpoint instead
2. remove the channel performance column
What's Changed
- opt: test for stream required channel by @looplj in https://github.com/looplj/axonhub/pull/681
- feat: auto backup to webdav, close #663 by @looplj in https://github.com/looplj/axonhub/pull/683
- feat: auto backup to data storage by @looplj in https://github.com/looplj/axonhub/pull/684
- feat: introduce watcher and live cache package by @looplj in https://github.com/looplj/axonhub/pull/685
- feat: add Antigravity provider with OAuth and endpoint fallback by @mcowger in https://github.com/looplj/axonhub/pull/670
- refactor: channel/api key migrate to live cache by @looplj in https://github.com/looplj/axonhub/pull/689
- fix: provider quota resovler deleted by mistaken by @looplj in https://github.com/looplj/axonhub/pull/690
- opt: migrate query channel override templates by @looplj in https://github.com/looplj/axonhub/pull/691
- opt: ui minor details by @looplj in https://github.com/looplj/axonhub/pull/692
- fix: remove constraint keywords from nested properties in antigravity sanitizer by @mcowger in https://github.com/looplj/axonhub/pull/693
- feat: fix missing pagination for gemini API model listing by @TundraWork in https://github.com/looplj/axonhub/pull/695
- Add citation and annotation support for completion responses by @mcowger in https://github.com/looplj/axonhub/pull/694
- refactor: migrate channel performance to probe by @looplj in https://github.com/looplj/axonhub/pull/697
- fix: dashboard should use calendar period, close #676 by @looplj in https://github.com/looplj/axonhub/pull/698
- refactor: unified base url normalization by @looplj in https://github.com/looplj/axonhub/pull/699
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/701
- refactor: Refactor date range picker with time selection support by @Libertyxxhappy in https://github.com/looplj/axonhub/pull/702
- feat: OpenAI image generation/edit API, close #347 by @looplj in https://github.com/looplj/axonhub/pull/705
- feat: gemini image generation via OpenAI image API by @looplj in https://github.com/looplj/axonhub/pull/707
- fix: allow reasoning_budget without reasoning_effort by @CoolZxp in https://github.com/looplj/axonhub/pull/706
- docs: update all docs by @looplj in https://github.com/looplj/axonhub/pull/709
- feat: support reasoning.summary and reasoning.generate_summary by @CoolZxp in https://github.com/looplj/axonhub/pull/710
- fix: require function_call arguments in output items by @CoolZxp in https://github.com/looplj/axonhub/pull/711
- doc: add opencode integration guideline by @looplj in https://github.com/looplj/axonhub/pull/712
- feat: change weight in place & add request curl button by @looplj in https://github.com/looplj/axonhub/pull/716
- feat: allow change developer to system options for all channels by @looplj in https://github.com/looplj/axonhub/pull/717
New Contributors
- @TundraWork made their first contribution in https://github.com/looplj/axonhub/pull/695
- @Libertyxxhappy made their first contribution in https://github.com/looplj/axonhub/pull/702
- @CoolZxp made their first contribution in https://github.com/looplj/axonhub/pull/706
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.8...v0.9.0

What's Changed
- fix: incorrect quota check condition for third party q provider by @looplj in https://github.com/looplj/axonhub/pull/679
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.7...v0.8.8

What's Changed
- feat: make requests table animation interval configurable via env var by @mcowger in https://github.com/looplj/axonhub/pull/660
- doc: fix tidb dsn missing parse time by @looplj in https://github.com/looplj/axonhub/pull/664
- Update: Claude Code OAuth and Enhanced Cloaking by @mcowger in https://github.com/looplj/axonhub/pull/640
- feat: add alternative paths for anthropic and gemini endpoints by @mcowger in https://github.com/looplj/axonhub/pull/666
- feat: add api key load balance strategy, close #655 by @looplj in https://github.com/looplj/axonhub/pull/665
- chore: sync model developers data by @github-actions[bot] in https://github.com/looplj/axonhub/pull/667
- feat: support claudecode/codex third party provider, close #662 by @looplj in https://github.com/looplj/axonhub/pull/671
- fix: codex prompt cache with session id by @looplj in https://github.com/looplj/axonhub/pull/673
- feat: add provider quota status tracking for Claude Code and Codex by @mcowger in https://github.com/looplj/axonhub/pull/669
- opt: skip third party codex/claudecode channel by @looplj in https://github.com/looplj/axonhub/pull/677
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.6...v0.8.7

What's Changed
- fix: channel test when strategy is circuit breaker by @looplj in https://github.com/looplj/axonhub/pull/659
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.5...v0.8.6

What's Changed
- fix: expand channel with many models cause ui bug by @looplj in https://github.com/looplj/axonhub/pull/621
- fix: should omit empty reasoning summary array by @looplj in https://github.com/looplj/axonhub/pull/622
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.3...v0.8.4

What's Changed
- chore: reorganize scripts and docs by @looplj in https://github.com/looplj/axonhub/pull/583
- refactor: fix retry pipeline & retry on channels instead of models by @looplj in https://github.com/looplj/axonhub/pull/585
- opt: channel test error message, close #582 by @looplj in https://github.com/looplj/axonhub/pull/586
- opt: customizable server host, close #587 by @looplj in https://github.com/looplj/axonhub/pull/588
- feat: codex support by @NicoOrz in https://github.com/looplj/axonhub/pull/578
- feat: introduce channel price setting and cost record by @looplj in https://github.com/looplj/axonhub/pull/591
- feat: add bailian dev-role toggle and merge consecutive tool_calls fo… by @x22x22 in https://github.com/looplj/axonhub/pull/590
- Feat/codex trace support by @x22x22 in https://github.com/looplj/axonhub/pull/592
- feat: anthropic write cache variant by @looplj in https://github.com/looplj/axonhub/pull/593
- refactor: flatten i18n keys by @looplj in https://github.com/looplj/axonhub/pull/594
- fix: bailian streaming output & codex compatibility by @x22x22 in https://github.com/looplj/axonhub/pull/595
- feat: api key quota limitation by @looplj in https://github.com/looplj/axonhub/pull/596
- opt: codex/claude code transformer by @looplj in https://github.com/looplj/axonhub/pull/597
- feat: backup channel model price by @looplj in https://github.com/looplj/axonhub/pull/598
- opt: gemini use parametersJsonSchema always by @looplj in https://github.com/looplj/axonhub/pull/599
- feat: channel model price auto match by @looplj in https://github.com/looplj/axonhub/pull/601
- refactor: oauth & codex auth by @looplj in https://github.com/looplj/axonhub/pull/604
- opt: disable auto complete for base url/api key, close #603 by @looplj in https://github.com/looplj/axonhub/pull/605
- opt: table page title & description by @looplj in https://github.com/looplj/axonhub/pull/606
- fix: input token cost calc by @looplj in https://github.com/looplj/axonhub/pull/607
- feat: auto refresh oauth token by @looplj in https://github.com/looplj/axonhub/pull/608
- fix: do not allow change email by user self by @looplj in https://github.com/looplj/axonhub/pull/611
- opt: channel row expand compatibility by @looplj in https://github.com/looplj/axonhub/pull/612
- chore: add render.yaml, close #610 by @looplj in https://github.com/looplj/axonhub/pull/613
- feat: channel stream capability by @looplj in https://github.com/looplj/axonhub/pull/614
- feat: add cost/token to the daily trending chart by @looplj in https://github.com/looplj/axonhub/pull/615
- chore: change dockerfile to use root, close #602 by @looplj in https://github.com/looplj/axonhub/pull/616
- feat: system timezone & codex/cc headers by @looplj in https://github.com/looplj/axonhub/pull/618
- opt: daily request stats chart by @looplj in https://github.com/looplj/axonhub/pull/619
- opt: remove duplicated request/usage log index by @looplj in https://github.com/looplj/axonhub/pull/620
New Contributors
- @NicoOrz made their first contribution in https://github.com/looplj/axonhub/pull/578
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.2...v0.8.3

BREAKING CHANGES
1. 删除了 user_roles 和 user_projects 的软删，需要手动将当前已删除的 user_roles 和 user_projects 记录删除
DELETE FROM user_projects WHERE deleted_at <> 0;
DELETE FROM user_roles WHERE deleted_at <> 0;
因为这个功能应该没什么人用，变动也不大，所以不发大版本了。
What's Changed
- opt: improve OpenAI API schema compatibility by @looplj in https://github.com/looplj/axonhub/pull/571
- opt: anthropic parallel tool call request compatibility, close #572 by @looplj in https://github.com/looplj/axonhub/pull/573
- Revert "fix: correct Gemini token count conversion to prevent negativ… by @looplj in https://github.com/looplj/axonhub/pull/575
- fix: remove soft delete to fix remove user from project not work by @looplj in https://github.com/looplj/axonhub/pull/579
- opt: response base url & transaction management by @looplj in https://github.com/looplj/axonhub/pull/580
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.1...v0.8.2

What's Changed
- fix: missing ui minor issues by @looplj in https://github.com/looplj/axonhub/pull/559
- feat: record request client ip, close #550 by @looplj in https://github.com/looplj/axonhub/pull/560
- opt: load trace/thread again when create with conflict, close #552 by @looplj in https://github.com/looplj/axonhub/pull/564
- opt: compatible with gemini query key auth, close #561 by @looplj in https://github.com/looplj/axonhub/pull/565
- opt: add open gpt-oss modles config, close #498 by @looplj in https://github.com/looplj/axonhub/pull/567
Full Changelog: https://github.com/looplj/axonhub/compare/v0.8.0...v0.8.1

BREAKING CHANGES
1. 增加了 models model id 唯一索引，如果启动失败，可以检查下是不是 model id 有冲突
What's Changed
- refactor(ui): extract channel models list into reusable component by @AAEE86 in https://github.com/looplj/axonhub/pull/482
- fix: start.ps1 should check the listen port only, close #483 by @looplj in https://github.com/looplj/axonhub/pull/484
- feat(ui): add i18n support for data table column headers by @AAEE86 in https://github.com/looplj/axonhub/pull/485
- refactor: move llm out of internal, and change LICENSE by @looplj in https://github.com/looplj/axonhub/pull/488
- feat(i18n): add translations for channel delete dialog by @AAEE86 in https://github.com/looplj/axonhub/pull/490
- fix: should stop retry if client canceled or deadline exceed, close #492 by @looplj in https://github.com/looplj/axonhub/pull/493
- build(docker): optimize build performance with layer caching and corepack by @AAEE86 in https://github.com/looplj/axonhub/pull/497
- opt: update request channel id and create new exection in txn, close #492 by @looplj in https://github.com/looplj/axonhub/pull/500
- fix: missing model id unique index by @looplj in https://github.com/looplj/axonhub/pull/501
- opt: channel and request table ui, close #489 by @looplj in https://github.com/looplj/axonhub/pull/502
- doc: add missing license by @looplj in https://github.com/looplj/axonhub/pull/503
- feat: backup and restore by @looplj in https://github.com/looplj/axonhub/pull/504
- opt: ui details improvement by @looplj in https://github.com/looplj/axonhub/pull/505
- feat: add model and channel executions to requests table by @looplj in https://github.com/looplj/axonhub/pull/506
- feat: add GitHub as a model provider by @WY-WY-W in https://github.com/looplj/axonhub/pull/509
- chore: format projects by @looplj in https://github.com/looplj/axonhub/pull/510
- feat: list models with channel restrictions by @looplj in https://github.com/looplj/axonhub/pull/512
- feat: auto disable channel setting by @looplj in https://github.com/looplj/axonhub/pull/513
- feat: allow hide mapped models, close #491 by @looplj in https://github.com/looplj/axonhub/pull/514
- opt: deduplicate candidate with same actual model by @looplj in https://github.com/looplj/axonhub/pull/517
- feat: add Claude Code transformer for channel integration by @mcowger in https://github.com/looplj/axonhub/pull/516
- opt: accent token color for dark theme by @looplj in https://github.com/looplj/axonhub/pull/518
- Convert all hardcoded Chinese error messages to i18n translation keys for dynamic language support. by @mcowger in https://github.com/looplj/axonhub/pull/521
- doc: update llm readme and api usage example by @looplj in https://github.com/looplj/axonhub/pull/522
- feat: add requests and tokens by API key dashboard metrics by @mcowger in https://github.com/looplj/axonhub/pull/523
- fix: 修复 OpenRouter/DeepSeek 思考内容丢失及 Content 字段 Null 值解析错误 by @billow96 in https://github.com/looplj/axonhub/pull/524
- opt(llm): override request headers by default when merge by @looplj in https://github.com/looplj/axonhub/pull/525
- opt: action column permission guard, and trace span type filter by @looplj in https://github.com/looplj/axonhub/pull/527
- fix: wrong gemini models response field, close #526 by @looplj in https://github.com/looplj/axonhub/pull/528
- opt: ui smooth animation by @looplj in https://github.com/looplj/axonhub/pull/529
- feat: add sorting by provider and status on channels page by @mcowger in https://github.com/looplj/axonhub/pull/530
- feat: channel health probe by @looplj in https://github.com/looplj/axonhub/pull/531
- feat: allow remove failed models in test dialog by @looplj in https://github.com/looplj/axonhub/pull/532
- feat: support json schema response format by @looplj in https://github.com/looplj/axonhub/pull/533
- opt: list models api object by @looplj in https://github.com/looplj/axonhub/pull/534
- opt: anthropic compatible for buggy channels, close #508 by @looplj in https://github.com/looplj/axonhub/pull/536

What's Changed
- doc: split claude code and codex integration by @looplj in https://github.com/looplj/axonhub/pull/466
- opt: retry to save with placeholder body, close #444 by @looplj in https://github.com/looplj/axonhub/pull/467
- style: increase provider list max height to 600px by @AAEE86 in https://github.com/looplj/axonhub/pull/468
- fix: model auto complete for model association, close #465 by @looplj in https://github.com/looplj/axonhub/pull/469
- feat: gemini allow input with pdf by @looplj in https://github.com/looplj/axonhub/pull/471
- fix: dashboard channels/models legend, close #470 by @looplj in https://github.com/looplj/axonhub/pull/472
- feat: add usage example to the api key view dialog, close #424 by @looplj in https://github.com/looplj/axonhub/pull/474
- feat: merge usage logs into requests table by @looplj in https://github.com/looplj/axonhub/pull/475
- feat: auto extract model prefix by @looplj in https://github.com/looplj/axonhub/pull/479
- fix: gemini optional parameters json schema fields by @looplj in https://github.com/looplj/axonhub/pull/480
- fix: bailian stream and infinite loop by @x22x22 in https://github.com/looplj/axonhub/pull/478
New Contributors
- @x22x22 made their first contribution in https://github.com/looplj/axonhub/pull/478
Full Changelog: https://github.com/looplj/axonhub/compare/v0.7.10...v0.7.11

What's Changed
- feat: add qwen model data by @looplj in https://github.com/looplj/axonhub/pull/436
- refactor(ui): flatten channel info layout in models association dialog by @AAEE86 in https://github.com/looplj/axonhub/pull/437
- feat: record request/request execution headers, close #421 by @looplj in https://github.com/looplj/axonhub/pull/438
- chore: cleanup console in fe code by @looplj in https://github.com/looplj/axonhub/pull/440
- fix: init and login page style by @looplj in https://github.com/looplj/axonhub/pull/442
- opt: unify scope select ui, close #439 by @looplj in https://github.com/looplj/axonhub/pull/446
- style(frontend): hide horizontal scrollbar on table containers by @AAEE86 in https://github.com/looplj/axonhub/pull/454
- feat: allow to configure channel transform options, close #441 by @looplj in https://github.com/looplj/axonhub/pull/453
- fix: fetched model list badge, close #451 by @looplj in https://github.com/looplj/axonhub/pull/456
- fix: fetch model with proxy, close #449 by @looplj in https://github.com/looplj/axonhub/pull/457
- fix: test gpt-5.x with max tokens, close #450 by @looplj in https://github.com/looplj/axonhub/pull/458
- fix: OpenAI prompt cache key field type by @looplj in https://github.com/looplj/axonhub/pull/459
- opt: retry with authorization header when fetch models, close #448 by @looplj in https://github.com/looplj/axonhub/pull/461
- opt: openai parallel_tool_calls compatibility, close #418 by @looplj in https://github.com/looplj/axonhub/pull/462
- fix: missing permission guard for bulk operations by @looplj in https://github.com/looplj/axonhub/pull/463
Full Changelog: https://github.com/looplj/axonhub/compare/v0.7.9...v0.7.10

What's Changed
- fix: system page dark theme background by @looplj in #415
- doc: reorganize readme and update to latest by @looplj in #416
- feat: sort channels/models by name by @looplj in #417
- opt: light theme brightness by @looplj in #422
Full Changelog: https://github.com/looplj/axonhub/compare/v0.7.7...v0.7.8