infigraph
An open-source code intelligence engine for developers and codebase owners that uses ASTs and graph-based retrieval to produce repo-scale code search, dependency tracing, and code understanding outputs.
Tool overview
Verdict: if you want a local-first, non-LLM foundation for code understanding and retrieval, infigraph looks worth tracking. But based on current evidence, it appears closer to a code intelligence backend for engineering teams and tool builders than a plug-and-play AI coding assistant. The available evidence is mostly the official GitHub repository description, which supports its positioning and architecture, but not broad real-world effectiveness; the sample is clearly limited.
In practical terms, the evidence says it is AST-powered, uses a graph database plus hybrid semantic search, supports analysis across 62 languages, and runs locally with zero LLM dependency. A more accurate analogy is an embeddable code indexing and relationship-analysis engine, not a Copilot-style autocomplete product and not just a generic vector database. It likely helps teams produce code search results, dependency maps, symbol-level understanding, and repository analysis capabilities.
On cost and adoption, there is no official pricing, hosted plan, or API fee information in the evidence.
This tool does not have related social references to display yet.