Cogni
PulseGate's liveness check found it on 3 Oct 2026; it has been in the index since 18 Aug 2026. How this is checked
Cogni is an MCP memory server that gives Claude, ChatGPT, Gemini, local models, and other agents connected long-term memory. Its entity-graph spreading activation retrieves linked facts across documents without using an LLM or GPU in the retrieval path.
Inferred · not functionally tested
Overview
6 featuresPurpose: Helping AI agents retrieve connected facts across documents instead of relying on isolated vector similarity matches.
Inferred · not functionally tested
Audience: AI developers and teams building LLM agents
Inferred · not functionally tested
Functions: rag, agents
Inferred · not functionally tested
Interfaces: API: unknown · MCP: indicated (inferred, not tested) · CLI: indicated (inferred, not tested) · Self-hosting: unknown
Recorded constraints: pricing: freemium · license: Proprietary · platforms: CLI, WEB · deployment: browser, cli, cloud_managed
Constraint provenance is unknown; confirm requirements with the publisher.
Record sources: getcogni.io. These links do not verify the individual claims.
Cogni is a Memory & skills project. Inferred · not functionally tested: It focuses on helping AI agents retrieve connected facts across documents instead of relying on isolated vector similarity matches. Inferred · not functionally tested: Cogni is a B2B product aimed at AI developers and teams building LLM agents. Basis unknown · not verified: A free plan is available. Basis unknown · not verified: It ships for the web and the command line.
Cogni builds and maintains Cogni. Inferred · not functionally tested: Among its 6 catalogued features are entity graph memory, spreading activation, and cross-vocabulary retrieval. Basis unknown · not verified: Catalogued interfaces include an MCP server.
Summary written by a language model from the project’s public pages.
Tasks: Inferred · not functionally tested
- Entity graph memory
- Spreading activation
- Cross-vocabulary retrieval
- MCP server
- Model-agnostic support
- Deterministic recall
Topics: Inferred · not functionally tested
Built with & integrations
- openai
- bgpt- in the HTML
- anthropic
- bclaude in the HTML
- Cloudflare
- cf-ray header · cf-cache-status header
- meta_llama
- llama- in the HTML
Trust & compliance
Indexing history
1What PulseGate has recorded for this listing
- Indexed18 Aug · 00:50 UTCShow HN: Cogni: MCP memory for LLMs, with no LLM in the retrieval path seen via Hacker News firehose (Algolia)Source: Hacker News firehose (Algolia) · Open
Frequently asked questions about Cogni
- What does Cogni do?
- Inferred · not functionally tested: Cogni focuses on helping AI agents retrieve connected facts across documents instead of relying on isolated vector similarity matches. It is catalogued under Memory & skills on PulseGate.
- Who is Cogni for?
- Inferred · not functionally tested: Cogni is a B2B product built for AI developers and teams building LLM agents.
- Is Cogni free?
- Basis unknown · not verified: Yes — there is a free tier, with paid plans for advanced use.
- What platforms does Cogni run on?
- Basis unknown · not verified: Cogni runs on the web and the command line.
- Is Cogni still active?
- PulseGate's liveness check found it on 3 Oct 2026.
- What are alternatives to Cogni?
- Similar projects tracked by PulseGate include Cognimemo, cogmemory-mcp, and Cog.Cognimemocogmemory-mcpCog
- Who develops Cogni?
- Cogni is developed by Cogni.
- Does Cogni have an API or integrations?
- Basis unknown · not verified: Yes — Cogni exposes an MCP server.
Similar projects
Closest matches by what these projects do