xerj
A local-first AI agent search engine supporting multiple file types and MCP servers.
- Region
- Overseas
- Pricing
- Free
- Open source
- Yes
- GitHub Stars
- ★ 16.8k
- Source
- GitHub
- Added
- 2026-10-27
- Last verified
- 2026-10-27
- GitHub
- github.com/xerj-org/xerj
Overview
XERJ is a local-first search engine designed for AI agents. It indexes files such as code, documents, logs, JSONL/CSV, and PDFs via a Rust binary, offering BM25 full-text, vector, and hybrid search capabilities. XERJ also supports code search, agent memory, and reference encoding, is compatible with the Elasticsearch API, and provides hundreds of reference memory corpora through Corpus Hub. Ideal for developers and teams needing efficient data search and processing.
Key features
- ▪Local-first search engine
- ▪Supports indexing of multiple file types
- ▪Compatible with Elasticsearch API
- ▪Built-in MCP server
- ▪Provides reference encoding
Use cases
Pros
- ＋Efficient local search
- ＋Supports multiple file types
- ＋Compatible with Elasticsearch API
- ＋Open source and easy to deploy
Limitations / notes
- －Requires internet to download related resources
- －Higher learning curve
Who it's for
This overview was compiled by AI from public sources and may contain inaccuracies — please refer to the official site.
FAQ
Is it free?
Yes, XERJ is an open-source project and can be used for free.
Do I need scientific internet access?
Partial resource downloads may require scientific internet access, but core functionality does not.
Does it support Chinese?
Yes, the README includes a simplified Chinese version.
Can it be used commercially?
Yes, XERJ is licensed under Apache-2.0, allowing commercial use.
Similar agents
Perplexity
FeaturedAI Search and Research Assistant
秘塔 AI 搜索
FeaturedAd-free Chinese AI search/research
Felo
Multilingual AI Search
Kimi 探索版
Deep search/research by Moonshot
纳米 AI 搜索
Multi-Agent AI Search by 360
OpenAI Deep Research
Multi-step Online Deep Research Agent
Something wrong? Let us know on the About page and we'll fix it.