Supermemory
A memory API for AI agents and apps that extracts facts and user profiles from chats, files and URLs, stores them in a vector-graph database and injects relevant memories back into the model context.
About Supermemory
Agents forget users and earlier work between sessions, and stitching together embeddings, vector stores and retrieval logic to fix that takes a lot of custom infrastructure.
Who it’s for
- Developers adding long-term, per-user memory to chatbots and agents
- Teams that want one hosted API for memory, document retrieval and search instead of running their own RAG stack
When to consider it
Consider Supermemory when you want per-user memory and document retrieval for an agent through one API, MCP server or plugin, with usage-based billing and an MIT-licensed open-source codebase.
Tradeoffs & limitations
- Usage is billed from a monthly credit balance: Free includes $5 of credits a month, Pro costs $19 a month ($20 included), Max $100 ($130 included) and Scale $399 ($600 included); memory ingestion is $5 per million SM tokens for plain text.
- Self-hosted deployments are offered only on the Scale and Enterprise plans; the GitHub repository is MIT licensed.
SOTA overview · Documentation-based assessment · Sources & review method
Updates
No updates shared yet.
Discussion
Newest firstAsk a question or share how you use Supermemory.
Keep it helpful. Community rules
Loading discussion…
Sources & review method
Documentation-based assessment · Nov 1, 2026 · Prepared with AI assistance; not a hands-on benchmark.
Checked by SOTA · AI-assisted documentation review. Selection advice is our assessment; verify current requirements for your deployment.
Import history & original evidence
Supermemory official website
A memory API for AI agents and apps that extracts facts and user profiles from chats, files and URLs, stores them in a vector-graph database and injects relevant memories back into the model context.
Based on official pages and announcements checked on 2026-11-01. No hands-on test, performance benchmark or popularity ranking is claimed.