osgrep and grepai

Both tools provide semantic search and code analysis capabilities for AI agents, but they compete on the same core functionality with slightly different trade-offs: osgrep emphasizes open-source accessibility with active downloads, while grepai prioritizes local-first execution and call graph analysis without external dependencies.

osgrep
64
Established
grepai
59
Established
Maintenance 10/25
Adoption 16/25
Maturity 22/25
Community 16/25
Maintenance 20/25
Adoption 10/25
Maturity 11/25
Community 18/25
Stars: 1,080
Forks: 66
Downloads: 369
Commits (30d): 0
Language: TypeScript
License: Apache-2.0
Stars: 1,443
Forks: 113
Downloads:
Commits (30d): 11
Language: C
License: MIT
No risk flags
No Package No Dependents

About osgrep

Ryandonofrio3/osgrep

Open Source Semantic Search for your AI Agent

Combines local ONNX embeddings with TreeSitter-based semantic chunking and call-graph tracing to provide concept-aware code search without external APIs. Integrates natively with Claude Code and Opencode plugins via a lightweight HTTP daemon (LanceDB-backed), automatically creating per-repository indexes while respecting `.gitignore` conventions. Role detection distinguishes orchestration logic from type definitions, enabling agent-optimized output with symbol maps and dependency analysis.

About grepai

yoanbernabeu/grepai

Semantic Search & Call Graphs for AI Agents (100% Local)

Builds vector embeddings locally using Ollama or LM Studio to enable semantic code search via natural language queries, then exposes results through an MCP server for direct integration with AI coding assistants like Claude Code and Cursor. Includes automatic call graph tracing to map function dependencies and a file watcher that keeps embeddings synchronized without manual reindexing.

Scores updated daily from GitHub, PyPI, and npm data. How scores work