Skip to content
Agents tracked: 442 Downloads (7d): 202M down 19.3% GitHub stars: 10.4M Releases (7d): 693 Agent status: 1 with issuesSince last limit reset: Claude 28d · Codex 1d Updated Oct 21, 2026

Cognee by Topoteretes

Open-source memory platform for AI agents that turns documents and conversations into a knowledge graph agents can search and reuse across sessions.

Agent tooling for cost, context and memory New · Added to AgentGid on Oct 20, 2026Open source · Apache-2.0Free to start
Price
Free (OSS)
free and open source
Popularity
#32 in Agent tooling
Gid Score 39 / 100 · #32 of 65 ranked agent tools
Trend
Steady
usage down 2.7% over 30 days
Development
Actively updated
latest 1.6.5, 5d ago
Community
32.6K GitHub stars
29.8K package downloads a week

Usage data: latest readings Oct 19 to Oct 21, 2026, last checked Oct 21, 06:35 UTC. Date of each number

What Cognee does

It is free and open source (Apache-2.0); you pay only for the models it calls.

Key facts

MakerTopoteretes
First releasedAugust 2023
LicenseApache-2.0
Main languagePython
Open issues and PRs971

Pricing

Open source under the Apache-2.0 licence; check the website for any paid hosted plans.

, checked Oct 20, 2026. Prices change often; confirm before you buy.

Compare prices of all agents

Quick answers

What's new in Cognee: October 2026 so far

Full monthly summary

October brought improvements to search accuracy and stability across Cognee, with new memory ingestion capabilities and better handling of vector data. The month focused on reliability fixes, restricted agent nesting, and optimized database operations.

Action needed: Removed temporalcognify configuration option — remove references from your configuration.
  • Natural-language searches now return matching database rows directly and skip unnecessary analysis steps.
  • Temporal searches now include boundary dates and rank date matches more accurately.
  • New Memify API endpoints let you convert content into memory entries programmatically.
  • Code docstrings are now stored as graph documents and linked to code symbols for better recall.

3 releases this month · summarised from the official release notes · all months

Get started with Cognee

PyPIpip install cognee

These commands use the packages tracked on this page. The vendor may recommend a different installer.

Compare Cognee with another agent

vs
# Agent Gid Score Downloads 7d 7d 30d Stars Latest release Price
1
OpenShellNVIDIA
77 101K down 14.1% up 381.4% 16.8K+103/day 0.1.5yesterday Free (OSS)
2
LiteLLM NewBerriAI
72 22.9M up 4.3% — 61.7K+79/day 1.104.5today Free (OSS)
3 69 — — — 22.8K+10/day — Free (OSS)
4
Langfuse NewLangfuse
64 9.5M down 17.1% up 17.7% 36.1K+48/day 4.66.0yesterday Free + $29/mo
5
Firecrawl NewFirecrawl
64 2M down 19.8% up 0.3% 194K+350/day 2.11.04mo ago Free + $16/mo
6
E2B NewE2B
60 3.2M down 14.5% up 21.9% 14.5K+21/day 2.11.2yesterday Free + $150/mo

Ratings from people who use Cognee

No ratings yet. Used Cognee? Rate it and say what it is good and bad at.

Cognee news

All AI agent news

  1. New on AgentGid AgentGid

    17 agents added to AgentGid

    Hindsight, Cherry Studio, Google Workspace CLI, Cognee, Hypit, GenOffice, Ego Lite, cc-haha, Hive, Moli, Vibe-Trading, MemOS and 5 more

  2. Release GitHub Releases

    Cognee 1.6.5 released

    Faster trace processing: traces are persisted and processed in fewer, larger operations for lower latency and fewer database writes.; Code-graph docstrings: function/class docstrings are stored as graph documents and linked to code symbols for better code…

  3. Release GitHub Releases

    Cognee 1.6.4 released

    Avoid returning the raw vector column during searches for LanceDB and pgvector adapters — reduces memory use and prevents failures.; Turso errors raised during database open now surface as OperationalError so calling code can handle them consistently…

What changed: recent Cognee releases

13 stable releases in the last 90 days · Monthly summary · Full changelog

1.6.5
  • Faster trace processing: traces are persisted and processed in fewer, larger operations for lower latency and fewer database writes.
  • Code-graph docstrings: function/class docstrings are stored as graph documents and linked to code symbols for better code recall and search.
  • New Memify API endpoints: convert content into memory entries via the API to make ingestion and downstream recall simpler.
  • Base64 embeddings support: request embeddings encoded as base64 where providers allow it to improve compatibility and reduce transport issues.
  • Security & connectors: daily CVE Slack alert, Google Calendar connector hardening, and improved OAuth/token sharing.
  • datamodel-code-generator requirement bumped: if you use a custom graphmodel generation step, you must have datamodel-code-generator >= 0.56.0. What to do: upgrade that dependency when generating…
1.6.4
  • Avoid returning the raw vector column during searches for LanceDB and pgvector adapters — reduces memory use and prevents failures.
  • Turso errors raised during database open now surface as OperationalError so calling code can handle them consistently.
  • Removed LanceDB background pruning to stop unexpected background work that could interfere with searches.
  • Added tests and harness fixes to prevent intermittent subprocess/session failures and improve stability.
  • Behavior change for LanceDB-backed datasets: automatic background pruning has been removed. If you relied on automatic pruning/compaction to reclaim storage space, you may need to run compaction…
  • No new end-user features in this release — this update focuses on bug fixes and reliability improvements.
Earlier releases
1.6.3
  • Temporal search windows now include their boundary dates and rank date matches more accurately.
  • Natural-language searches return matching database rows in the response payload and skip 'turn analysis' to avoid unrelated context.
  • Stronger ingestion validation rejects malformed node sets and fixes Notion block rendering (no depth caps).
  • Agents restrictions tightened: agents can no longer create nested agents; parent/agent dataset access behavior corrected.
  • Operational fixes: Docker images now include git and a DLT dependency is pinned to avoid import failures.
  • DLT dependency pinned below 1.31: The DLT package used for some ingestion paths is now pinned to versions older than 1.31 to avoid import errors. Action: if you manage your own environment, pin…
1.6.2
  • Embedding fixes prevent silent truncation and size chunks by each model's token limit so your content is embedded intact and consistently.
  • Slack integration: import and continuously sync conversation history from Slack into Cognee.
  • Skills moved to a new .agents/skills location with Windows linking support and recovery when linking fails.
  • Better errors and faster failure on unrecoverable embedding errors — you’ll see real causes instead of opaque messages.
  • CLI and docs refreshed (including database migration/commands and Docker quickstart auth fixes).
  • Skills path change: Built-in and packaged skills are now under .agents/skills (moved from earlier locations). If you maintain custom scripts, automation, build hooks, or CI that reference the old…
1.6.1
  • New Google Drive OAuth connector plus Gmail and shared-drive sync — easier, safer import from Google workspaces.
  • Streaming /visualize/json responses in chunks so large graph views load without exhausting memory.
  • GLiNER installer overhauled: installs off the event loop, auto-installs CPU Torch on first use, and reports install telemetry and errors.
  • Chunks now carry document external metadata and that metadata is surfaced in hybrid retrieval (combined semantic+lexical search).
  • dlt is now a core dependency: The CSV loader no longer gates imports on dlt; dlt has been promoted to a core dependency. If you install Cognee from source or manage dependencies manually, ensure…
  • Google Drive OAuth connector: Adds a full OAuth-based connector for Google Drive that lets you authorize and sync files from your Google account or shared drives. Why it matters: simplifies…
1.6.0
  • Keyless-first experience: local model downloads are announced and many flows now run cleanly with no LLM key.
  • Pipeline robustness: pipeline runs are stamped at start and recover reliably after crashes, preserving completed documents.
  • Model wiring: image and audio transcription models are passed through end-to-end and the embedding model is recorded per dataset to ensure consistent embeddings.
  • Telemetry & privacy: dataset names are fingerprinted before leaving the host and error telemetry now reports full routes for better diagnostics.
  • Developer tooling: a new cognee-mcp client/server and CLI inclusions for Enola improve integration and testing workflows.
  • GLiNER removed from default Docker image — The default Docker image no longer bundles GLiNER tooling. If you relied on GLiNER being present in the standard image, you must add it to your image or…

Cognee popularity and usage data

Cognee was downloaded 29.8K times from PyPI in the week ending Oct 19, 2026, down 0.3% on the week before and down 2.7% over the last 30 days compared with the 30 before. The GitHub repository has 32.6K stars, 1,896 of them added in the last 30 days. The latest stable release is 1.6.5, published Oct 16, 2026; there were 5 stable releases in the past 30 days.

Downloads per dayPyPI, 7-day average 4,256/day
GitHub starsrunning total 32.6K
Website rankTranco list; a lower rank means more traffic #678,681

Download the raw daily series: cognee.csv

Gid Score breakdown

Adoption33
Community63
Attention—
Momentum47

Each component is scored 0 to 100 from public signals; a dash means no data for it. Methodology.

How fresh these numbers are

NumberLatest readingLast checked
Package downloadsOct 19, 2026Oct 21, 06:31 UTC
GitHub starsOct 21, 2026Oct 21, 06:32 UTC
Website rank (Tranco)Oct 20, 2026Oct 21, 06:35 UTC

We fetch every source once a day. Registries publish downloads and page views a day or two late, so their latest reading trails the check. A reading more than 4 days old is marked out of date; the Gid Score still uses it until a newer one arrives.

Cognee FAQ

How much does Cognee cost?

Open source under the Apache-2.0 licence; check the website for any paid hosted plans.

Is Cognee open source?

Yes. Cognee is open source, released under the Apache-2.0 license.

How popular is Cognee?

Cognee was downloaded 29.8K times from PyPI in the week ending Oct 19, 2026, down 0.3% on the week before and down 2.7% over the last 30 days compared with the 30 before. The GitHub repository has 32.6K stars, 1,896 of them added in the last 30 days. The latest stable release is 1.6.5, published Oct 16, 2026; there were 5 stable releases in the past 30 days.

What is the latest version of Cognee?

The latest stable release we track is 1.6.5, published on Oct 16, 2026.

What are the alternatives to Cognee?

The closest alternatives in the same category by Gid Score are OpenShell, LiteLLM, NVIDIA NemoClaw, Langfuse.

For the makers of Cognee badges, corrections

Add a live badge to your README or site:

AgentGid: #32 in Agent tooling Gid Score: 39/100 downloads: 29.8K/week ▼3%

[![AgentGid: #32 in Agent tooling](https://agentgid.com/badges/cognee-rank.svg)](https://agentgid.com/agents/cognee/)
[![Gid Score: 39/100](https://agentgid.com/badges/cognee-score.svg)](https://agentgid.com/agents/cognee/)
[![downloads: 29.8K/week ▼3%](https://agentgid.com/badges/cognee-downloads.svg)](https://agentgid.com/agents/cognee/)

Something wrong or outdated on this page? Send a correction with a link to the official source. More for makers.

Where these numbers come from

PyPI: cognee. GitHub: topoteretes/cognee. Site rank: cognee.ai on the Tranco list. See data sources for how each one is collected.

Product facts and pricing were checked against: cognee.ai, github.com, pypi.org.