Peer-centric memory built on four primitives: Workspaces (app-level isolation), Peers (any long-lived entity — user, agent, group, project, or idea), Sessions (many-to-many interaction threads between peers), and Messages (data units that trigger reasoning). Honcho 3.0 (Jan 2026) split reasoning into three stages run by the deriver worker process: ingestion reasoning extracts deductive (explicit) and inductive (inferred) conclusions in parallel on write; an async Dreaming Agent handles lower-priority background work (summarization, peer-card generation, surprisal-prioritized reprocessing) off the request path; a query-time Dialectic Agent answers chat() calls with an agentic search-and-synthesize loop across five reasoning depths (minimal to max). Configurable LLM backends (OpenAI, Anthropic, Gemini); queries target this derived conclusion/representation layer rather than raw message history.
Deep diveHoncho deep dive: how theory-of-mind user modeling actually works in productionRead →- Storage
- Managed cloud at api.honcho.dev, or self-hosted FastAPI server. Persistent layer: PostgreSQL + pgvector (default vector backend); Turbopuffer and LanceDB supported as alternatives. Optional Redis caching for hot paths.
- Retrieval
- Hybrid BM25 + vector search scoped to workspace, session, or peer level. Context endpoint aggregates messages, conclusions, and session summaries within a caller-specified token budget. Representation endpoint delivers low-latency static snapshots of derived peer insights. Chat API provides reasoning-grounded natural-language responses against the live peer model.
- Self-host
- Self-host: heavy
- License
- AGPL-3.0
- Pricing
- $2/M tokens ingested (storage + background reasoning included); retrieval unlimited; Reasoning API $0.001–$0.50/query by depth tier; $100 free credits on signup; startup program: $1,000 credits + 12 months subsidized pricing; enterprise custom · Freemium
- GitHub stars
- 6,557
- Last release
- —
- Last commit
- 2026-08-06
- First catalogued
- 2026-06-28
Strengths
- Theory-of-mind user modeling: derives conclusions about beliefs, preferences, and contradictions rather than replaying raw chat history
- Peer-centric multi-entity model: any entity (user, agent, group, project, idea) is a Peer; tracks what one peer knows about another across many-to-many sessions
- Broad integration surface: remote MCP (Claude Code, Cursor, Windsurf, VS Code, Zed, Cline, Codex, Goose), Python + TypeScript SDKs, REST API, and a first-party Claude Code plugin
- Hybrid retrieval: BM25 + vector search combined with a context endpoint that aggregates conclusions within caller-specified token budgets
- Tiered reasoning cost: $2/M for ingestion; per-query Reasoning API ($0.001–$0.50) lets callers pay only for the inference depth they need
Watch out
- Background reasoning fires LLM calls on ingestion (and dreaming/dialectic calls on top) even when self-hosted — adds latency and cost regardless of hosting mode; AGPL-3.0 restricts commercial embedding without a commercial license
- Git tags exist (e.g. v3.0.12) but there are no published GitHub Releases or changelog notes — pinning a version means tracking tags yourself, not reading release notes
- Heavy self-host footprint: PostgreSQL + pgvector + Redis + deriver worker process + an LLM backend all required; multi-service orchestration is non-trivial for small teams
- No first-party endpoint to wipe a peer's accumulated memory (a 'peer delete' endpoint is on the roadmap as of Apr 2026); the documented workaround is versioning peer names or deleting directly in Postgres, and headline benchmark scores (LongMemEval 90.4, LoCoMo 89.9) are self-reported without a disclosed embedder
Best for
- Personalization that must model a user's evolving beliefs, preferences, and contradictions over many sessions (theory-of-mind, not fact recall)
- Multi-agent or multi-party systems where what peer A knows about peer B — or what an AI agent knows about a human — needs to persist and update
- Coding-assistant memory (Claude Code plugin, Cursor MCP) where context must survive across sessions, directories, and projects
How it integrates
Benchmark results
| Benchmark | Value | Backbone | Trust | Source |
|---|---|---|---|---|
| longmemeval | 90.4 accuracy | — | Self-reported | Honcho (Plastic Labs) ↗ |
| locomo | 89.9 accuracy | — | Self-reported | Honcho (Plastic Labs) ↗ |
Sources
- Honcho 3 announcement re-read to re-substantiate the carded pricing — $2 per million tokens ingested, Chat API reasoning tiers minimal $0.001 / low $0.01 / medium $0.05 / high $0.10 / max $0.50 per query, context() not charged, dreaming tasks free for now. Note honcho.dev/pricing and the docs pricing path both 404 as of 2026-08-01, so the $100 signup credits and startup-program terms on this card could not be re-confirmed from a live vendor page (vendor)
- Monthly card re-verification — GitHub API repo metadata: 6,356 stars, license AGPL-3.0 confirmed, archived=false, last commit on the default branch 2026-07-29 (third-party)
- https://honcho.dev/ (vendor)
- https://github.com/plastic-labs/honcho (vendor)
- Honcho MCP integration — remote-http at mcp.honcho.dev; Claude Code, Cursor, Windsurf, VS Code, Cline, Zed, Codex, Goose clients (vendor)
- Honcho Claude Code plugin — /plugin marketplace add plastic-labs/claude-honcho; per-directory/branch/chat session strategies; linked-host cross-tool reads (vendor)
- Honcho README — peer-centric primitives, deriver pipeline, hybrid BM25+vector retrieval, vector backend options (pgvector/Turbopuffer/LanceDB), SDK packages (vendor)
- Honcho review — theory-of-mind, peer-centric (human/agent/idea) personalization (third-party)
- GitHub API — 5,669 stars, pushed 2026-06-30, AGPL-3.0 (vendor)
- Announcing Honcho 3 (2026-01-26) — splits reasoning into ingestion (deductive/inductive conclusions), async Dreaming Agent, and query-time Dialectic Agent; $2/M token pricing replaces per-message billing (vendor)
- Chat (Dialectic) endpoint docs — five reasoning levels (minimal/low/medium/high/max) trading latency for synthesis depth (vendor)
- Honcho evals page — self-reported LongMemEval 90.4 ('LongMem S'), LoCoMo 89.9, BEAM 100K 0.630; backbone/embedder not disclosed on the page (vendor)
- 'Wiping agent memory' issue — no native peer-delete endpoint as of close (Apr 28, 2026); founder confirms a delete endpoint is on the roadmap, workaround is versioned peer names or direct DB deletes (vendor)
- OpenCode, OpenClaw, and Hermes integrations confirmed live (README + docs) (vendor)
- GitHub API re-check — license still AGPL-3.0, not archived, pushed 2026-07-23T18:22Z; still no published GitHub Releases (confirmed via /releases endpoint returning empty); latest tag v3.0.12 (third-party)
- Honcho API spec re-check — confirms no DELETE /peers/{peer_id} endpoint exists (only session-scoped peer removal); the 'no peer-delete endpoint' watchOut still holds (third-party)
- Honcho pricing re-check — unchanged: $2.00/M tokens (storage + reasoning), Reasoning API $0.001 (minimal) to $0.50/query (max), $100 free signup credits, startup program ($1,000 credits + 12mo subsidized pricing for companies raising <$5M) (vendor)
Last verified 2026-08-01 · updated by refresh-framework-cards