v2.16.0

trycompai/compv2.16.0Feb 16, 2025by Leizhenpeng
AI Summary

Major update adding DeepSeek API integration with streaming reasoning support, MCP protocol support, and expanded AI model options including SiliconFlow provider.
Key Highlights

DeepSeek API integration with streaming reasoning content support
MCP (Model Context Protocol) support added
New AI models: o3-mini, gemini-2.0-flash-thinking-exp, deepseek-reasoner
SiliconFlow provider with VLM (Vision Language Model) support
Vision model runtime configuration
DeepSeek model provider with separate API URL and key support
Enhanced model icons and SVG support
New Features

DeepSeek API integration
MCP protocol support
New AI models (o3-mini, gemini-2.0-flash-thinking, deepseek-reasoner)
SiliconFlow provider with VLM support
Vision model runtime configuration
DeepSeek streaming reasoning content
Shortcut key to clear context
Model leak prevention with unique IDs
Full Release Notes

## What's Changed
* chore(deps-dev): bump @testing-library/jest-dom from 6.6.2 to 6.6.3 by @dependabot in https://github.com/ChatGPTNextWeb/NextChat/pull/5807
* fix: use current session id to trigger rerender by @Dogtiti in https://github.com/ChatGPTNextWeb/NextChat/pull/5812
* 更新视觉模型匹配关键词 by @LovelyGuYiMeng in https://github.com/ChatGPTNextWeb/NextChat/pull/5815
* artifacts support svg by @OPChips in https://github.com/ChatGPTNextWeb/NextChat/pull/5816
* Fix gemini summary by @DDMeaqua in https://github.com/ChatGPTNextWeb/NextChat/pull/5819
* fix: 代码块嵌入小代码块时渲染错误 by @OPChips in https://github.com/ChatGPTNextWeb/NextChat/pull/5827
* feat: support more user-friendly scrolling by @Sherlocksuper in https://github.com/ChatGPTNextWeb/NextChat/pull/5821
* update new model for gpt-4o and gemini-exp by @code-october in https://github.com/ChatGPTNextWeb/NextChat/pull/5861
* Fix o1 by @code-october in https://github.com/ChatGPTNextWeb/NextChat/pull/5850
* 🎨 style(setting): Place custom-model's input a separated row. by @frostime in https://github.com/ChatGPTNextWeb/NextChat/pull/5879
* 面具“以文搜图”改成“AI文生图”，微调提示让图片生成更稳定无水印 by @InitialXKO in https://github.com/ChatGPTNextWeb/NextChat/pull/5936
* chore(deps-dev): bump @testing-library/react from 16.0.1 to 16.1.0 by @dependabot in https://github.com/ChatGPTNextWeb/NextChat/pull/5940
* Update google models to add gemini-2.0 by @fengzai6 in https://github.com/ChatGPTNextWeb/NextChat/pull/5932
* add gemini-exp-1206, gemini-2.0-flash-thinking-exp-1219 by @dupl in https://github.com/ChatGPTNextWeb/NextChat/pull/5959
* feat: runtime configuration of vision-capable models by @Yiming3 in https://github.com/ChatGPTNextWeb/NextChat/pull/5919
* Update README.md by @Leizhenpeng in https://github.com/ChatGPTNextWeb/NextChat/pull/5976
* fix model leak issue by @code-october in https://github.com/ChatGPTNextWeb/NextChat/pull/5883
* fix: failed unit test by @Dogtiti in https://github.com/ChatGPTNextWeb/NextChat/pull/5994
* Feature/cogview by @Dogtiti in https://github.com/ChatGPTNextWeb/NextChat/pull/5996
* fix: correct the URL in plugin store address by @xixici in https://github.com/ChatGPTNextWeb/NextChat/pull/5993
* Fix issue #5964: Prevents character loss in gemini-2.0-flash-thinking-exp-1219 responses by @zmhuanf in https://github.com/ChatGPTNextWeb/NextChat/pull/5965
* Use i18n for DISCOVERY by @fishshi in https://github.com/ChatGPTNextWeb/NextChat/pull/5920
* feature: support glm-4v by @Dogtiti in https://github.com/ChatGPTNextWeb/NextChat/pull/5997
* Use regular expressions to make the code more concise. by @dupl in https://github.com/ChatGPTNextWeb/NextChat/pull/5998
* since #5984, add DeepSeek as a new ModelProvider (with deepseek-chat&deepseek-coder models), so that user can use openai and deepseek at same time with different api url & key by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/5989
* docs: add DEEPSEEK_API_KEY and DEEPSEEK_URL in README by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/6001
* fix issue #6009  add setting items for deepseek by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/6016
* 修复 VISION_MDOELS 在 docker 运行阶段不生效的问题 by @code-october in https://github.com/ChatGPTNextWeb/NextChat/pull/6010
* feat: #5422 快捷键清除上下文 by @DDMeaqua in https://github.com/ChatGPTNextWeb/NextChat/pull/5468
* Correct the typos in user-manual-cn.md by @dupl in https://github.com/ChatGPTNextWeb/NextChat/pull/6034
* fix: prevent message sync between forked sessions by generating unique IDs by @lvguanjun in https://github.com/ChatGPTNextWeb/NextChat/pull/6033
* Support MCP( WIP) by @Leizhenpeng in https://github.com/ChatGPTNextWeb/NextChat/pull/5974
* fix: add missing useMobileScreen hook import by @Kadxy in https://github.com/ChatGPTNextWeb/NextChat/pull/6075
* Update README_CN.md by @Leizhenpeng in https://github.com/ChatGPTNextWeb/NextChat/pull/6082
* fix: missing mcp_config.json files required for building by @Kadxy in https://github.com/ChatGPTNextWeb/NextChat/pull/6084
* Bug fix  (about mcpPrompt) and add deepseek-reasoner by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/6087
* Support DeepSeek API streaming reasoning content by @Kadxy in https://github.com/ChatGPTNextWeb/NextChat/pull/6116
* fix: revert unintended changes by @Kadxy in https://github.com/ChatGPTNextWeb/NextChat/pull/6118
* Update README.md by @Leizhenpeng in https://github.com/ChatGPTNextWeb/NextChat/pull/6119
* feat(model): add support for OpenAI o3-mini model by @Kadxy in https://github.com/ChatGPTNextWeb/NextChat/pull/6121
* add gemini-2.0-flash-thinking-exp, gemini-2.0-flash-thinking-exp-01-21 by @dupl in https://github.com/ChatGPTNextWeb/NextChat/pull/6127
* fix: fix isModelNotavailableInServer logic for bytedance models by @zcong1993 in https://github.com/ChatGPTNextWeb/NextChat/pull/6134
* temporary fix for o3-mini by @RetiredQQ in https://github.com/ChatGPTNextWeb/NextChat/pull/6132
* chore: add knowledge cut off dates for o1 and o3 by @MonadMonAmi in https://github.com/ChatGPTNextWeb/NextChat/pull/6129
* New provider SiliconFlow and Its Latest DeekSeek Models by @jackalcooper in https://github.com/ChatGPTNextWeb/NextChat/pull/6141
* feat: add more llm icons by @Eric-2369 in https://github.com/ChatGPTNextWeb/NextChat/pull/6144
* Fix: 补充 env 中硅基流动的环境变量；追加硅基流动 2 个支持的付费模型 by @xiexin12138 in https://github.com/ChatGPTNextWeb/NextChat/pull/6168
* update the lastest Gemini models by @dupl in https://github.com/ChatGPTNextWeb/NextChat/pull/6158
* Fix: Set consistent fill color for OpenAI/MoonShot/Grok SVG to prevent color inversion in dark mode by @ZhangYichi-ZYc in https://github.com/ChatGPTNextWeb/NextChat/pull/6164
* Larger timeout for SiliconFlow by @jackalcooper in https://github.com/ChatGPTNextWeb/NextChat/pull/6185
* Fix formatting of reasoning model on SiliconFlow by @jackalcooper in https://github.com/ChatGPTNextWeb/NextChat/pull/6186
* Add other Xai model by @itsevin in https://github.com/ChatGPTNextWeb/NextChat/pull/6175
* fix several bugs by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/6172
* Update LICENSE by @Leizhenpeng in https://github.com/ChatGPTNextWeb/NextChat/pull/6188
* Support VLM on SiliconFlow by @jackalcooper in https://github.com/ChatGPTNextWeb/NextChat/pull/6194
* fix avatar for export message preview and saved image by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/6189
* Model listing of SiliconFlow by @jackalcooper in https://github.com/ChatGPTNextWeb/NextChat/pull/6193
* Fix model icon on SiliconFlow by @jackalcooper in https://github.com/ChatGPTNextWeb/NextChat/pull/6190
* add 3 type of reasoning_content support (+deepseek-r1@OpenAI @Alibaba @ByteDance), parse <think></think> from SSE by @bestsanmao in https://github.com/ChatGPTNextWeb/NextChat/pull/6204

## New Contributors
* @LovelyGuYiMeng made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5815
* @Sherlocksuper made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5821
* @dupl made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5959
* @Yiming3 made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5919
* @xixici made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5993
* @zmhuanf made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5965
* @fishshi made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5920
* @bestsanmao made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/5989
* @lvguanjun made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6033
* @Kadxy made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6075
* @zcong1993 made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6134
* @RetiredQQ made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6132
* @MonadMonAmi made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6129
* @jackalcooper made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6141
* @xiexin12138 made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6168
* @itsevin made their first contribution in https://github.com/ChatGPTNextWeb/NextChat/pull/6175

**Full Changelog**: https://github.com/ChatGPTNextWeb/NextChat/compare/v2.15.8...v2.16.0