Diffbot

https://www.diffbot.com/

Auto-built knowledge graph from web crawl. 1T+ facts, 10B+ entities. GraphRAG-fine-tuned model based on open-source Llama 3.3.

At a glance

Type
Auto-extracted KG from web crawl
Tier
T1
Created
2008
Latest release
not applicable — not OSS
License
not applicable — not OSS
GitHub
not applicable — no GitHub repo
Pricing
SaaS
Funding
$12.5M total (Seed + Series A 2016)

Taxonomy

storage
graph
retrieval
graph-traversal
persistence
long-term
update
extraction
unit
fact
governance
opaque
conflict
llm-arbitrate

When to use

Optimised for: relationship modeling + reasoning + governance over pure vector

Anti-fit: not for purely-vector or simple-RAG use cases (graph adds setup cost)

Pros & cons

Pros

Pre-built knowledge graph of the web (KGoW) — you don't build the graph, you query someone else's; saves enormous extraction cost.

Cons

Quality is bounded by Diffbot's extraction; no schema control; not appropriate for proprietary internal data.

Claims & capabilities

10B+ entities and 1T+ structured facts crawled from public web; 81% accuracy on FreshQA (Google factuality benchmark) — beats ChatGPT and Gemini; 70.36% MMLU-Pro; customers include Cisco, DuckDuckGo, Snapchat, Adobe, AOL, eBay, Microsoft

Technical surface

API surface
REST + GraphQL (DQL Knowledge Graph; Enhance, Extract APIs)
Backend storage
custom
Deployment
Managed-only
Embedding model
locked
Multi-tenancy
hard-isolation
MCP
no first-party MCP adapter published as of 2026-05; community connectors may exist.
A2A
no Google A2A (Agent2Agent) integration documented as of 2026-05.
OpenTelemetry
no first-party OpenTelemetry exporter documented; standard logs/metrics typically available.

Compare Diffbot with…

Similar systems

Other knowledge-graph platforms in the catalog, ranked by inbound references.

  • Neo4j T1

    Property graph DB with Cypher; native vector search; Aura Agent + MCP server for graph-as-memory. $100M GenAI investment in 2025.

  • Amazon Neptune Analytics T1

    Vector index on graph nodes queryable via openCypher. Mem0 integration GA 2025; Cognee integration for agentic RAG. Combines semantic recall with multi-hop traversal in one managed service.

  • AllegroGraph (Franz) T1

    RDF triple/quad store with RDFS++ / OWL reasoning. v8.4 (May 2025) added an NLQ interface.

  • Apache AGE T2

    openCypher over Postgres. Pairs with pgvector for graph + vector hybrid retrieval. Azure Database for PostgreSQL ships AGE with ai_extension LLM functions for entity extraction directly in SQL.

  • ArangoDB T1

    HybridGraphRAG combines vector search, graph traversal, full-text in one AQL query. ArangoGraphML for ML pipelines; LangChain integration.

  • Dgraph T2

    Vector indexing on any node. Google Gen AI Toolbox integration; LangChain agent orchestration.

Related systems

References (1)

  • Meta Llama 4 family depends on at runtime — AG-fine-tuned model based on open-source Llama 3.3.

Row last verified 2026-05-14. Catalog data is CC-BY-4.0 — see how to read this.