DocuMemory
DocuMemory distils AI chat sessions into searchable engineering decisions and prompt-ready context. Engineers use it to carry useful project knowledge into later sessions with connected AI clients.
One of 71 tools in Coding
Who it's for
- AI-first engineers using Claude Code, ChatGPT, Cursor, or Windsurf who want to retain decisions from earlier sessions.
- Engineering teams that want a shared documentation repository with project access restricted to selected people.
- Engineers who want a local-first or self-hosted memory setup using their own Postgres database and AI keys.
Not the right fit if…
- People who need ongoing free use: the free allowance is 15 uses once per account, after which continued use requires a paid plan.
- Teams with fewer than two seats: the Team plan has a two-seat minimum.
How it fits your workflow
Create an access token in Settings and add DocuMemory as an MCP connection in a supported AI client. The page shows a configuration using the DocuMemory MCP endpoint and a bearer token.
Ask the client to save a conversation. DocuMemory distils it into structured memory, then hybrid search can retrieve relevant facts and provide compact context for a later session.
Pricing
Free: The free trial includes 15 uses once per account, with no credit card required. Distillation and Enhance with AI share this allowance.
Free trial
US $0
- · 15 uses, once per account
- · No credit card required
- · Explore the workspace and connect an AI client
- · Distillation and Enhance with AI share the free allowance
Monthly
US $12/month, billed monthly
- · Continue after the free uses
- · Documents
- · AI integrations
- · Private uploads
Annual
US $10/month, US $120 billed annually
- · Continue after the free uses
- · Documents
- · AI integrations
- · Private uploads
Team
US $15 per seat per month, or US $150 per seat billed annually; 2-seat minimum
- · One shared documentation repository for the team
- · Every member gets a seat
- · Share projects with everyone or restrict them to specific people
- · No free trial; billing starts when the organization is set up
Prices checked on Oct 19, 2026 from the vendor's site. They can change; confirm before you buy.
Key features
- Chat-to-document distillation
- Capture an AI conversation and distil it into structured engineering documentation that retains useful facts and reasoning while removing noise.
- MCP-native capture and recall
- Connect an AI client through MCP to capture, search, and recall engineering knowledge from the client.
- Hybrid search
- Combines semantic and full-text search using reciprocal rank fusion (RRF), supporting both conceptual queries and exact identifiers.
- Prompt-ready context snippets
- Produces compact context blocks intended to give a later AI session relevant facts without requiring the full chat transcript.
- Local-first and self-hosted options
- The page describes self-hosting with the user's own Postgres database and AI keys, including use with a local model.
- Shared team documentation
- The Team plan provides a shared documentation repository, with projects that can be shared with everyone or restricted to specific people.
Works with
- Claude Code
- ChatGPT
- Cursor
- Windsurf
- MCP clients
Limitations to know
- The free allowance is limited to 15 uses once per account, and distillation and Enhance with AI share that allowance.
- The Team plan requires at least two seats and has no free trial; billing starts when the organization is set up.
Alternatives to consider
-
Atlaso
Choose it if you want a memory layer described as automatically capturing and recalling context across multiple AI tools, rather than saving a conversation through DocuMemory's capture workflow.
-
Recalld
Choose it if you need memory that stores both conversations and documents and reconciles new facts with existing memory through an API or MCP server.
-
Rutter
Choose it if you want an open-source memory tool that stores decision summaries and hashes of cited Markdown notes in a folder you own.
Getting started
- Sign up and use the free allowance to explore the workspace and connect an AI client.
- Create an access token in Settings.
- Configure the MCP connection in your AI client.
- Ask the client to save a conversation.