v0.6.8

What's Changed

• fix minimax empty log by @ye4293 in #1560
• ci: use codecov to upload coverage report by @LeoQuote in #1583
• refactor: replace hardcoded string with ctxkey constant by @igophper in #1579
• chore(deps): bump golang.org/x/image from 0.16.0 to 0.18.0 by @dependabot in #1568
• chore: Add support for Claude Sonnet 3.5 by @shaoyun in #1567
• refactor: abusing goroutines and channel by @zijiren233 in #1561
• feat: support spark4.0 ultra by @lihangfu in #1569
• feat: support Spark4.0 Ultra by @shijilin0116 in #1575
• fix: fix size not support during image generation by @igophper in #1564
• docs: add related projects by @da730 in #1562
• ci: skip archive, upload directly by @LeoQuote in #1586
• feat: add support for Claude 3 tool use (function calling) by @dirname in #1587
• refactor: init db by @igophper in #1590
• feat: support one or more log file by @linzeliang1222 in #1400
• feat: support smtp without auth by @LeoQuote in #1101
• feat(ui): show available models for air theme by @LeoQuote in #1595
• feat: support test specific model by @WqyJh in #1600
• feat: cloudflare support native openai api by @zijiren233 in #1596
• fix: use musl libc by @zijiren233 in #1597
• fix: update readme docs by @peterwillcn in #1599
• feat: add initial root access token by @peterwillcn in #1598
• feat: add Novita AI as model provider by @jasonhp in #1609
• feat: refactor AwsClaude to Aws to support both llama3 and claude by @WqyJh in #1601
• fix: duplicate [DONE] by @WqyJh in #1629
• fix: impl cloudflare worker ai gateway by @zijiren233 in #1617
• fix: 修复4v请求下text参数缺少引发的nil by @ye4293 in #1634
• feat: support claude and gemini in vertex ai by @vaayne in #1621
• feat: add field stop in GeneralOpenAIRequest by @f1ynn-zhan9 in #1637
• chore(deps): bump google.golang.org/grpc from 1.64.0 to 1.64.1 by @dependabot in #1641
• fix: rate limit can be zero by @zijiren233 in #1643
• feat: fast build linux/arm64 frontend by @zijiren233 in #1645
• feat: vertexai support proxy url(example: cloudflare ai gateway) and fix some vertexai bug by @zijiren233 in #1642
• feat: extend anthropic header to support 8k context windows by @Laisky in #1654
• fix: support embedding models for doubao by @igophper in #1662
• feat: fast build linux/arm64 frontend by @zijiren233 in #1663
• feat: support gpt-4o mini by @HynoR in #1665
• feat: add Proxy channel type and relay mode by @Laisky in #1678

New Contributors

• @LeoQuote made their first contribution in #1583
• @shaoyun made their first contribution in #1567
• @zijiren233 made their first contribution in #1561
• @da730 made their first contribution in #1562
• @linzeliang1222 made their first contribution in #1400
• @peterwillcn made their first contribution in #1599
• @jasonhp made their first contribution in #1609
• @vaayne made their first contribution in #1621
• @f1ynn-zhan9 made their first contribution in #1637
• @HynoR made their first contribution in #1665

Full Changelog: v0.6.7...v0.6.8

v0.6.7

What's Changed

• feat: support gpt-4o by @WqyJh in #1431
• feat: berry theme update & bug fix by @MartialBE in #1471
• fix: fix gpt-4o token encoding by @Kxzuir in #1446
• chore: print user id when relay error happened by @ye4293 in #1447
• fix: fix react-apexcharts version by @fatwang2 in #1465
• fix: add prefixes to image models to solve the problem of duplicate models by @mo2g in #1469
• feat: generate default token after register by @carey036 in #1401
• feat: Support gemini embeddings (text-embedding-004,embedding-001) by @mxdlzg in #1475
• feat: add support for bytedance's doubao by @klb3713 in #1438
• feat: post all messages for cloudflare by @fxsome in #1515
• fix: xunfei interface call 4001 error by @lynnssb in #1499
• fix: support for Spark Lite model by @laxtiz in #1526
• fix: berry theme channel edit settings key error by @MartialBE in #1496
• fix: fix default token not created in some cases by @jinjianming in #1510
• fix: Gemini streaming return being truncated by FinishReason. by @mxdlzg in #1477
• feat: support tencent v3 api by @lihangfu in #1542
• fix: fix sparkdesk function call by @shijilin0116 in #1544

New Contributors

• @fatwang2 made their first contribution in #1465
• @carey036 made their first contribution in #1401
• @klb3713 made their first contribution in #1438
• @fxsome made their first contribution in #1515
• @lynnssb made their first contribution in #1499
• @laxtiz made their first contribution in #1526
• @jinjianming made their first contribution in #1510
• @lihangfu made their first contribution in #1542
• @shijilin0116 made their first contribution in #1544

Full Changelog: v0.6.6...v0.6.7

v0.6.6

What's Changed

• chore(deps): bump golang.org/x/net from 0.17.0 to 0.23.0 by @dependabot in #1335
• feat: support aws bedrockruntime claude3 by @Laisky in #1328
• fix: logger race by @Laisky in #1339
• feat: groq support Llama3 now by @tylinux in #1333
• feat: add function and tools support for Gemini by @mxdlzg in #1358
• fix: fix default theme blank screen when edit channel again by @tylinux in #1363
• feat: add cohere support by @ye4293 in #1355
• fix: refactor Gemini adaptor to support streaming content generation by @mxdlzg in #1382
• feat: support for the ollama vision model by @nongmo677 in #1376
• feat: use mapped model name to test by @tylinux in #1370
• fix: claude stream response parse by @WqyJh in #1334
• fix: fix zhipu embedding error when input is array but not string by @caixinjiang in #1306
• fix: fix getPreConsumedQuota by @plusye in #1312

New Contributors

• @tylinux made their first contribution in #1333
• @mxdlzg made their first contribution in #1358
• @nongmo677 made their first contribution in #1376
• @caixinjiang made their first contribution in #1306
• @plusye made their first contribution in #1312

Full Changelog: v0.6.5...v0.6.6

v0.6.5

What's Changed

• fix: update model-ratio.go 修正文心计费模型名称 by @manjieqi in #1241
• feat: add embedding-2 support for zhipu by @cdredfox in #1273
• feat: update baidu model name & ratio by @manjieqi in #1253
• feat: update baidu model name & ratio by @manjieqi in #1277
• feat: support Ali stable-diffusion-xl and wanx-v1 model by @mo2g in #1240
• fix: fix Lark icon button style by @AI-ASS in #1279
• feat: berry theme update & bug fix by @MartialBE in #1282
• chore: resolve the issue of onclick event scope for custom Lark button by @AI-ASS in #1281
• feat: support function call for ali (close #1242)
• feat: support top_k for claude (close #1239)
• feat: initial function call support for xunfei
• feat: able to set model limitation for token (close #178)
• feat: /v1/models now only return available models
• feat: add subnet validation (#1275)
• feat: support feishu login now
• fix: do not detect quota field in error message (close #1276)
• feat: show token info when quota is not enough (close #1274)
• feat: able to change gemini version (close #1211)
• feat: add gpt-4-turbo support (close #1304)

New Contributors

• @manjieqi made their first contribution in #1241
• @cdredfox made their first contribution in #1273
• @mo2g made their first contribution in #1240

Full Changelog: v0.6.4...v0.6.5

v0.6.4

What's Changed

• feat: 移除 azure model 的 TrimSuffix by @lgxisbb in #1193
• feat: 支持 ollama 的 embedding 接口 by @xietong in #1221
• fix: 修复了讯飞的错误处理
• feat: 现在调用 /api/token 会返回创建的 token 的信息了
• fix: 修复了 audio 相关接口的错误时未返还 quota 的问题
• chore: 更新了主题 berry 的文案
• fix: 修复了自动填充的模型列表不正确的问题
• fix: 修复了在未启用缓存的情况下重试时依然使用最高优先级的问题

New Contributors

• @lgxisbb made their first contribution in #1193
• @xietong made their first contribution in #1221

Full Changelog: v0.6.3...v0.6.4

v0.6.3

What's Changed

• fix: 修复berry主题下令牌在编辑后点击新建弹窗值的初始化问题 by @yooyui in #1165
• fix: fix ratio of gpt-3.5-turbo (close #1011) by @mrhaoji in #1163
• fix: change Moonshot value to 25 by @zu1k in #1158
• chore(deps): bump github.com/jackc/pgx/v5 from 5.3.1 to 5.5.4 by @dependabot in #1157
• ZhiPuAI compatibility with the temperature topP parameter (0.0, 1) by @majian159 in #1091
• 零一万物的模型价格错误了 by @ye4293 in #1182
• feat: add user list sorting and pagination enhancements (#77, #602) by @mrhaoji in #1178
• fix: fix panel cards style(修复penal卡片高度不一致) by @shuiRong in #1171
• feat: support Azure OpenAI TTS. by @techotaku in #1177
• feat: update default API version for Azure OpenAI by @techotaku in #994
• docs: update README by @mrhaoji in #1186

New Contributors

• @zu1k made their first contribution in #1158
• @majian159 made their first contribution in #1091
• @shuiRong made their first contribution in #1171

Full Changelog: v0.6.2...v0.6.3

v0.6.2

What's Changed

• 支持 Claude 3 系列模型
• 现在渠道编辑页面可以直接填充与本渠道相关的全部模型了
• 支持了 Groq 渠道
• 删除了之前为国产大模型不支持 system 提示词时的兼容措施
• 优化了 stream 模式下的错误检测
• 现在渠道编辑页面可以直接搜索渠道类型与模型名称了
• 现在用户被删除、封禁后将实时生效
• 现在可以根据错误率自动禁用渠道了，需要手动开启，详见 README 中 METRIC 相关环境变量部分
• 修复了开启 Turnstile 后用用户无法修改密码的问题
• 现在支持通过 Message Pusher 将报警信息推送到各种 App 中了
• 现在 One API 将自动为你补齐缺失的模型的倍率
• 现在支持在渠道编辑页面传入大 key 了
• 现在日志只会保存到一个文件中，除非重新启动
• 现在会在 panic 时输出更多的信息以供 debug
• 现在当缓存中用户额度过低时，会强制刷新缓存
• 支持了 Ollama 渠道
• 为百度渠道添加了更多的 Embeddable 模型支持
• fix: model name typo by @sparanoid in #1109
• 修复百度模型中URL的字符串判断bug； by @Jguobao in #1143
• 修改AMA问天为BotGem(官方改名) by @afafw in #934
• Update OtherSetting.js by @yooyui in #1148
• chore(deps): bump google.golang.org/protobuf from 1.30.0 to 1.33.0 by @dependabot in #1145
• feat: add support for private registry in docker-compose.yml by @warjiang in #1103

New Contributors

• @sparanoid made their first contribution in #1109
• @Jguobao made their first contribution in #1143
• @afafw made their first contribution in #934
• @yooyui made their first contribution in #1148
• @warjiang made their first contribution in #1103

Full Changelog: v0.6.1...v0.6.2

v0.6.1

What's Changed

1. 修复了版本号为空的问题；
2. 支持了百川的模型；
3. 支持了 MiniMax 的模型；
4. 支持了 ChatGLM-4；
5. 修复了 Azure 渠道的测试功能；
6. 支持了 Mistral 的模型；
7. 支持了在重试时切换到其他优先级的渠道；
8. 修复了 Ali 的模型在 relay 到时候部分参数被 ignore 掉的问题；
9. 当进行渠道测试的时候，如果 gpt-3.5-turbo 不在模型列表中，则使用模型列表中的第一个模型进行测试；

New Contributors

• @momomobinx made their first contribution in #1046

Full Changelog: v0.6.0...v0.6.1

v0.6.0

What's Changed

• fix: remove useless wrong index by @Calcium-Ion in #916
• refactor: refactor relay part by @songquanpeng in #935
• fix: aff not effective with berry by @MartialBE in #937
• refactor: refactor relay part by @songquanpeng in #957
• fix: fix primary chat button in berry theme by @yongman in #951
• feat: sync models with OpenAI by @mrhaoji in #971
• chore: update code implementation by @Laisky in #978

New Contributors

• @songquanpeng made their first contribution in #935
• @yongman made their first contribution in #951

Full Changelog: v0.5.11...v0.6.0

v0.5.11

What's Changed

• fix: gpt-4 vision 传入 base64 字符串图片时解析错误 by @MartialBE in #858
• feat: 改进阿里模型 stream 模式 & 使其支持联网搜索能力 by @moondie in #856
• chore(deps): bump golang.org/x/crypto from 0.14.0 to 0.17.0 by @dependabot in #840
• fix: fix SearchUsers not working if using PostgreSQL by @ImSingee in #778
• fix: openai response should contains model by @Laisky in #841
• feat: able to change gemini safety setting by @tisfeng in #867
• fix: base 64 encoded format support of gemini-pro-vision for field im… by @liuzl in #878
• feat: add support for davinci-002 and babbage-002 by @Tailen in #888
• fix: fix button copywriting by @seven-yu in #880
• fix: support base64 encoded image_url for gemini-pro-vision by @Laisky in #872
• fix: fix token validation exception handling by @seven-yu in #901
• 一个现代化的web UI界面 by @MartialBE in #860
• fix: fix some issues with berry by @MartialBE in #913
• fix: fix bugs with theme berry by @MartialBE in #931

New Contributors

• @moondie made their first contribution in #856
• @Laisky made their first contribution in #841
• @tisfeng made their first contribution in #867
• @liuzl made their first contribution in #878
• @Tailen made their first contribution in #888
• @seven-yu made their first contribution in #880

Full Changelog: v0.5.10...v0.5.11