steel-dev
★ 7,645steel-browser
🔥 Open Source Browser API for AI Agents & Apps. Steel Browser is a batteries-included browser sandbox that lets you automate the web without worrying about infrastructure.
AI 频道大模型应用 
★ 7,117
★ 7,022
开源大模型应用与框架
这是范围最广的一组：聊天界面、SDK 和框架、提示词与评测工具，以及各种基于大语言模型的应用。很多项目可以接入多家服务商的模型，有些还能在你自己的硬件上运行模型。
怎么选
打算用在产品里的话，先确认三件事：许可证——有些热门项目限制商用；是否支持你正在用的模型服务商；以及它是否还在维护。每个项目页上的「秒懂」卡片会根据 GitHub 数据回答许可证和维护这两个问题。
Star 最多的项目
第 241–264 个，共 864 个tailcallhq
★ 7,629forgecode
AI enabled pair programmer for Claude, GPT, O Series, Grok, Deepseek, Gemini and 300+ models
flyteorg
★ 7,487flyte
该项目围绕「Dynamic, resilient AI orchestration. Coordinate data, models, and compute as you build AI workflows.」构建，适用于实际场景的开源实践，提供可复用的工具链与集成方式。
open-compass
★ 7,442opencompass
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.
traceloop
★ 7,430openllmetry
Open-source observability for your GenAI or LLM application, based on OpenTelemetry
guardrails-ai
★ 7,418guardrails
Adding guardrails to large language models.
apache
★ 7,397hertzbeat
An AI-powered next-generation open source real-time observability system.
arcee-ai
★ 7,350mergekit
Tools for merging pretrained large language models.
zhouxiaoka
★ 7,335autoclip
AutoClip : AI-powered video clipping and highlight generation · 一款智能高光提取与剪辑的二创工具
luhengshiwo
★ 7,329LLMForEverybody
每个人都能看懂的大模型知识分享，LLMs春/秋招大模型面试前必看，让你和面试官侃侃而谈
strands-agents
★ 7,241harness-sdk
构建代理线束并对其进行端到端控制。用于 Python 和 TypeScript 中的生产 AI 代理的开源 SDK - 任何模型、任何云。
Zipstack
★ 7,238unstract
LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows
zai-org
★ 7,190GLM-5
GLM-5: From Vibe Coding to Agentic Engineering
plastic-labs
★ 7,180honcho
Memory library for building stateful agents
JerryZLiu
★ 7,125Dayflow
The automatic work journal/time tracker. Privately turns your screen into a timeline of what you actually accomplished. Open-source and local-first.
中文LLaMA-2 & Alpaca-2大模型二期项目 + 64K超长上下文模型 (Chinese LLaMA-2 & Alpaca-2 LLMs with 64K long context models)
lance-format
★ 7,076lance
多模式 AI 的开放 Lakehouse 格式。只需 2 行代码即可从 Parquet 进行转换，以实现速度提高 100 倍的随机访问、向量索引和数据版本控制。与 Pandas、DuckDB、Polars、Pyarrow 和 PyTorch 兼容，即将推出更多集成。
datajuicer
★ 7,047data-juicer
Data processing for and with foundation models! 🍎 🍋 🌽 ➡️ ➡️🍸 🍹 🍷
TalkToFigma: MCP integration between AI Agent (Cursor, Claude Code, Codex) and Figma, allowing Agentic AI to communicate with Figma for reading designs and modifying them programmatically.
lintsinghua
★ 7,020DeepAudit
DeepAudit 是一个开源多智能体代码审计平台，提供漏洞挖掘与 PoC 验证能力，支持 Ollama 私有部署，并可一键生成审计报告，适合快速上手。
opensquilla
★ 7,012opensquilla
OpenSquilla，相同预算、更高智能密度的代币高效AI代理。
mishushakov
★ 6,927llm-scraper
Turn any webpage into structured data using LLMs
open-multi-agent
★ 6,922open-multi-agent
具有动态工作流程的 TypeScript AI 代理编排框架。描述目标，而不是图表：协调员在运行时规划任务 DAG，并在任何 LLM（Claude、ChatGPT、Gemini、DeepSeek 或本地模型）上运行它。
Gentleman-Programming
★ 6,829gentle-ai
Gentle-AI configures the AI coding agents you already use: Claude Code, Cursor, OpenCode, Codex, Pi, and more. Choose persistent memory, Spec-Driven Development, curated skills, MCP servers, personas, and optional bounded review. Open source, no agent lock-in.