memnode
Memnode gives your AI agents a persistent graph memory so they stop forgetting between conversations and stop wasting tokens replaying context. Instead of re-stuffing every prior fact into the prompt, your agent calls Memnode to record what it learns and recall what's relevant later — answers come back in tens of milliseconds with a short "why" trace explaining each supporting memory. Unlike…
1 subscribers
8.5/10 popularity
312 ms avg latency
33% success rate
7 endpoints
The in-depth APIMemo review for this API hasn't been published yet —
the data below comes straight from the public marketplace listing.
memnode endpoints
| Method | Endpoint | Description |
|---|---|---|
| memory | ||
| POST |
queryMemory /v1/query |
Exact, structured lookup over the graph. Use when you already know the entity label (e.g. `Service`, `Person`) and want to filter by an identity property — not for fuzzy retrieval. |
| POST |
recallMemory /v1/recall |
Ask a natural-language question and get a synthesized answer plus the supporting memories, scored confidence, and a short "why" trace. Use this whenever the agent is about to… |
| POST |
forgetMemory /v1/forget |
Marks the node as archived so it stops appearing in recalls. Use `force: true` to archive even when the node has dependents (graph edges that derive from it). |
| POST |
recordMemory /v1/record |
Persist something the agent learned. Provide either a free-form `observation` (most common) **or** a `entities` + `relationships` set for structured graph writes. Memories are… |
| feedback | ||
| POST |
rateMemory /v1/feedback |
After a recall, tell Memnode which memory actually answered the question. Feedback adjusts ranking for future recalls in your workspace. |
| introspection | ||
| GET |
getMemoryLineage /v1/lineage/{node_id} |
|
| GET |
getHealth /healthz |
Unauthenticated. Returns `{"ok":true}` if the data plane is alive. |
memnode pricing
| Plan | Price | Rate limit | Quotas |
|---|---|---|---|
| BASIC | Free | — |
|
