# 多源 AI 新闻摘要 (过去 48 小时)
采集时间: 2026-03-27T13:23:22.866458+00:00
## 🔥 GitHub Trending 热门 AI 仓库
- 总计采集: 45 条 (GitHub Trending 42, GitHub Releases 3)
## 📋 按质量评分排序（星标越多越热门）:
1. **[sgl-project/sglang: SGLang — 面向大语言模型和多模态模型的高性能服务部署框架](https://github.com/sgl-project/sglang)**
- ⭐ 今日新增星标: 31 | 总计: 25,110
- SGLang is a high-performance serving framework for large language models and multimodal models.
2. **[yamadashy/repomix: 📦 Repomix — 强大工具，将整个代码仓库打包为单个 AI 友好文件，适合将代码喂给大语言模型使用](https://github.com/yamadashy/repomix)**
- ⭐ 今日新增星标: 36 | 总计: 22,683
- 📦 Repomix is a powerful tool that packs your entire repository into a single, AI-friendly file. Perfect for when you need to feed your codebase to Large Language Models (LLMs) or other AI tools like C
3. **[NVIDIA/TensorRT-LLM: NVIDIA TensorRT-LLM — 为用户提供易用的 Python API 定义大语言模型，支持最先进优化在 NVIDIA GPU 上高效推理](https://github.com/NVIDIA/TensorRT-LLM)**
- ⭐ 今日新增星标: 14 | 总计: 13,201
- TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT
4. **[0xPlaygrounds/rig: ⚙️🦀 在 Rust 中构建模块化可扩展的大语言模型应用](https://github.com/0xPlaygrounds/rig)**
- ⭐ 今日新增星标: 10 | 总计: 6,672
- ⚙️🦀 Build modular and scalable LLM Applications in Rust
5. **[thu-pacman/chitsu (赤兔) — 高性能大语言模型推理框架，专注于效率、灵活性和可用性](https://github.com/thu-pacman/chitu)**
- ⭐ 今日新增星标: 11 | 总计: 4,282
- High-performance inference framework for large language models, focusing on efficiency, flexibility, and availability.
6. **[skyzh/tiny-llm — 系统工程师学习苹果芯片上大语言模型推理服务部署教程：从零构建 vLLM + 通义千问](https://github.com/skyzh/tiny-llm)**
- ⭐ 今日新增星标: 12 | 总计: 4,035
- A course of learning LLM inference serving on Apple Silicon for systems engineers: build a tiny vLLM + Qwen.
7. **[bitsandbytes-foundation/bitsandbytes — 通过 k-bit 量化让大语言模型在 PyTorch 上更易访问](https://github.com/bitsandbytes-foundation/bitsandbytes)**
- ⭐ 今日新增星标: 5 | 总计: 8,082
- Accessible large language models via k-bit quantization for PyTorch.
8. **[open-compass/opencompass — OpenCompass 大语言模型评测平台，支持超过 100+ 数据集评测多种模型 (Llama3, Mistral, InternLM2, GPT-4, LLaMa2, 通义千问, GLM, Claude 等)](https://github.com/open-compass/opencompass)**
- ⭐ 今日新增星标: 7 | 总计: 6,803
- OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets.
9. **[ModelTC/LightLLM — LightLLM 是基于 Python 的大语言模型推理和服务部署框架，以轻量设计、易扩展和高性能著称](https://github.com/ModelTC/LightLLM)**
- ⭐ 今日新增星标: 4 | 总计: 3,978
- LightLLM is a Python-based LLM (Large Language Model) inference and serving framework, notable for its lightweight design, easy scalability, and high-speed performance.
10. **[stanford-crfm/helm — 斯坦福 CRFM 创建的语言模型整体评测框架，开源、可复现、透明](https://github.com/stanford-crfm/helm)**
- ⭐ 今日新增星标: 2 | 总计: 2,722
- Holistic Evaluation of Language Models (HELM) is an open source Python framework created by the Center for Research on Foundation Models (CRFM) at Stanford for holistic, reproducible and transparent e
11. **[llm-jp/awesome-japanese-llm: 日本語LLMまとめ - Overview of Japanese LLMs](https://github.com/llm-jp/awesome-japanese-llm)**
- ⭐ 今日新增星标: 1 | 总计: 1,357
- 日本語LLMまとめ - Overview of Japanese LLMs
12. **[trpc-group/trpc-agent-go: trpc-agent-go is a powerful Go framework for building intelligent agent systems using large language models (LLMs) and tools.](https://github.com/trpc-group/trpc-agent-go)**
- ⭐ 今日新增星标: 3 | 总计: 1,041
- trpc-agent-go is a powerful Go framework for building intelligent agent systems using large language models (LLMs) and tools.
13. **[Optima-CityU/LLM4AD: LLM4AD: A Platform for Algorithm Design with Large Language Model](https://github.com/Optima-CityU/LLM4AD)**
- ⭐ 今日新增星标: 1 | 总计: 671
- LLM4AD: A Platform for Algorithm Design with Large Language Model
14. **[agentscope-ai/Trinity-RFT: Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (LLM).](https://github.com/agentscope-ai/Trinity-RFT)**
- ⭐ 今日新增星标: 2 | 总计: 570
- Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (LLM).
15. **[sdcb/chats: A powerful and flexible frontend & AI gateway for large language models, supporting 21+ mainstream AI model providers.](https://github.com/sdcb/chats)**
- ⭐ 今日新增星标: 1 | 总计: 561
- A powerful and flexible frontend & AI gateway for large language models, supporting 21+ mainstream AI model providers.
16. **[CherryHQ/cherry-studio — AI 生产力工作室，支持智能聊天、自主智能体和 300+ 助手，统一访问前沿大语言模型](https://github.com/CherryHQ/cherry-studio)**
- ⭐ 今日新增星标: 63 | 总计: 42,432
- AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs
17. **[CN-Syndra/ClawLink — AI 智能体社交网络 — 让你的 AI 直接和其他人的 AI 对话，支持自主智能体间通信、协商和协作，完全由所有者控制](https://github.com/CN-Syndra/ClawLink)**
- ⭐ 今日新增星标: 44 | 总计: 308
- An AI Agent Social Network — Let your AI (Claw) talk to other people's AIs directly. Autonomous agent-to-agent communication, negotiation, and collaboration, with full owner control.
18. **[openclaw v2026.3.24](https://github.com/openclaw/openclaw/releases/tag/v2026.3.24)**
19. **[tang-vu/ContribAI — 自主 AI 智能体为开源做贡献：自动发现仓库、分析代码、生成修复、提交 PR](https://github.com/tang-vu/ContribAI)**
- ⭐ 今日新增星标: 19 | 总计: 172
- Autonomous AI agent that contributes to open source — discovers repos, analyzes code, generates fixes, and submits PRs
20. **[stakpak/agent — 自动交付你的代码，开源智能体 24/7 运行在你的机器上保持应用运行](https://github.com/stakpak/agent)**
- ⭐ 今日新增星标: 3 | 总计: 1,198
- Ship your code, on autopilot. An open source agent that lives on your machines 24/7 and keeps your apps running. 🦀
21. **[jd-opensource/JoySafeter: 🚀 JoySafeter 企业级 AI 智能体平台 — 不止聊天，通过可视化编排构建、运行、测试和追踪自主智能体团队](https://github.com/jd-opensource/JoySafeter)**
- ⭐ 今日新增星标: 3 | 总计: 226
- 🚀 JoySafeter: An enterprise AI Agent Platform—Not just chatting. building、running、testing, and tracing autonomous Agent Teams with visual orchestration...
22. **[ertugrulakben/cashclaw — 将你的 AI 智能体变成赚钱机器，50+ HYRVE API 端点，作业轮询守护进程，自动接受模式，v1.6.1](https://github.com/ertugrulakben/cashclaw)**
- ⭐ 今日新增星标: 4 | 总计: 131
- Turn your AI agent into a money-making machine. 50+ HYRVE API endpoints, job polling daemon, auto-accept mode. v1.6.1
23. **[OpenCowAI/opencow — 一个任务一个智能体，交付完成。开源任务驱动自主 AI 智能体平台，为每个任务分配自主智能体](https://github.com/OpenCowAI/opencow)**
- ⭐ 今日新增星标: 8 | 总计: 118
- One task, one agent, delivered. The open-source platform for task-driven autonomous AI agents.OpenCow assigns an autonomous AI agent to every task — features, campaigns, reports, audits — and deliver
24. **[ythx-101/openclaw-qa — OpenClaw Q&A 社区 — AI 智能体记忆系统、多智能体架构、进化系统、具身AI | 龙虾茶馆 🦞](https://github.com/ythx-101/openclaw-qa)**
- ⭐ 今日新增星标: 2 | 总计: 76
- OpenClaw Q&A 社区 — AI Agent 记忆系统、多Agent架构、进化系统、具身AI | 龙虾茶馆 🦞
25. **[TONresistor/teleton-agent: Teleton: Autonomous AI Agent for Telegram & TON Blockchain](https://github.com/TONresistor/teleton-agent)**
- ⭐ 今日新增星标: 1 | 总计: 66
- Teleton: Autonomous AI Agent for Telegram & TON Blockchain
26. **[hoangsonww/Agentic-AI-Pipeline: 🦾 A production‑ready research outreach AI agent that plans, discovers, reasons, uses tools, auto‑builds cited briefings, and drafts tailored emails with tool‑chaining, memory, tests, and turnkey Docke](https://github.com/hoangsonww/Agentic-AI-Pipeline)**
- 🦾 A production‑ready research outreach AI agent that plans, discovers, reasons, uses tools, auto‑builds cited briefings, and drafts tailored emails with tool‑chaining, memory, tests, and turnkey Docke
## 📊 统计
- 大语言模型 (LLM): 15 条
- AI 智能体 (AI Agent): 11 条
- 总计: 44 条去重后