Search files semantically using embeddings without external APIs. For developers who need fast, local semantic grep across codebases and documents.
io.github.ArkNill/embgrep
Local install
STDIO
No auth required
How models use it and what it is built for.
Search files semantically using embeddings without external APIs. For developers who need fast, local semantic grep across codebases and documents.
Local install — runs as a subprocess.
Where to find authoritative docs and source for Embgrep.
Paste any of these into Agent Studio after connecting Embgrep.
Common questions about connecting and running Embgrep.
Does embgrep require internet or external API keys?
No. Embgrep runs entirely locally with zero external service dependencies, making it suitable for offline use and private codebases.
How do I install and run embgrep?
Install via `uvx embgrep@0.1.0`. It communicates over stdio, so it integrates directly with MCP-compatible clients and tools.
What file types does embgrep search?
The registry metadata does not specify supported file types. Check the project repository or documentation for details on which formats are indexed.
How is embgrep different from grep or ripgrep?
Embgrep uses embeddings for semantic search, so it finds conceptually similar code even if keywords don't match exactly—unlike regex-based tools that require literal text matches.
What are the performance and storage requirements?
The registry does not document performance characteristics or disk space needs. Refer to the project documentation for benchmarks and hardware recommendations.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent Studio