OpenGraph AI
Machine-readable overview. This page is intentionally plain text, no images, no animations, no scripts required to read the content.
# Summary
OpenGraph AI is an open-source semantic graph layer that extracts, links, and reasons across multimodal data (tables, text, images, audio, video). It exposes a queryable knowledge graph that AI agents can traverse as a tool.
# Capabilities
- ingest: PDF, CSV, JSON, MP3/MP4, JPG/PNG, URLs, object storage
- extract: entities, relations, concepts, citations across modalities
- resolve: cross-modal entity resolution into a single graph
- query: natural language, graph queries, or as an agent tool
# Interfaces
- MCP server (Claude and other MCP-compatible runtimes)
- CLI
- REST API
- Agent skill (skill.md)
# Key routes
- / — homepage
- /how-it-works — pipeline overview
- /opengraph-image — image-to-graph workflow
- /opengraph-text — text-to-graph workflow
- /opengraph-table — table-to-graph workflow
- /use-cases — use case index
- /playground — interactive graph explorer
- /resources — docs and references
- /about — project info
- /auth — sign in / sign up
# License
Open source. Source: https://github.com/OpenGraphAI/opengraph-ai
# For humans
If you are a human and reached this page by mistake, use the button at the top to return to the main site.