需先复制 .env.example 并填写必需的环境变量,否则无法启动;依赖 Bun 作为包管理器,需提前安装;本地运行前必须先执行数据库迁移 bun db:deploy
agentset
agentset-aiMIT活跃维护开箱即用的开源 RAG 平台,内置引用溯源、深度研究和分区隔离,支持 22+ 文件格式与 MCP 服务,帮团队快速搭出可落地的检索增强应用。
The open-source RAG platform: built-in citations, deep research, 22+ file formats, partitions, MCP server, and more.
cp .env.example .env
bun install
bun db:deploy
bun dev:webREADME 未说明
开源 RAG 平台,覆盖数据摄取、向量索引、评测到托管部署的全流程
开箱即用的开源 RAG 平台,内置引用溯源、深度研究和分区隔离,支持 22+ 文件格式与 MCP 服务,帮团队快速搭出可落地的检索增强应用。
核心能力
- 内置摄取、分块、嵌入与检索的完整 RAG 流水线
- 模型无关,可自选 LLM、嵌入模型和向量数据库
- 聊天调试台支持消息编辑与引用溯源
- 提供 API、类型化 SDK 和 OpenAPI 规范
- 内置多租户与带预览链接、自定义域名的生产托管
容易踩的坑
暂无 release 数据(可能没有发版,或尚未同步)。
用户点评
登录用户的真实使用记录,可按版本和硬件筛选还没有人点评。第一个说出真实体验的人最有价值。
Features
- Turnkey RAG: ingestion, chunking, embeddings, and retrieval
- Model agnostic: works with your choice of LLM, embeddings, and vector DB
- Chat playground with message editing and citations
- Production hosting with preview links and custom domains
- API + typed SDKs, OpenAPI spec
- Built-in multi-tenancy
- Built with TypeScript, Next.js, AI SDK, Prisma, Supabase, and Trigger.dev
Getting Started
Agentset Cloud
The fastest way to get started with Agentset. Generous free tier with 1,000 pages and 10,000 retrievals. No credit card required.
- Sign up: https://app.agentset.ai/login
Self-host Agentset
Follow our complete guide: https://docs.agentset.ai/open-source/prerequisites
Quick Start (Local Development)
# 1) Copy env and fill required values
cp .env.example .env
# 2) Install dependencies
bun install
# 3) Run database migrations (from the repo root)
bun db:deploy
# 4) Start the app
bun dev:web
Useful scripts:
bun db:studio– open Prisma Studiobun dev:web– run only the web app
Star Us
If you find Agentset useful, please give the repo a star — it helps a lot!
Contributing
We <3 contributions big and small. Feel free to:
- Open an issue for bugs and feature ideas
- Submit a PR with focused changes
- Improve docs or examples
Not sure where to start? Check existing issues: https://github.com/agentset-ai/agentset/issues
License
MIT :)
Made with ❤️ by the Agentset team.