Agent MemoryUnclaimed

Cognee

Open-source AI memory platform that builds knowledge graphs to help agents retrieve relevant context across sessions

Also known as
cognee

Available worldwide · Popular in: EU, US

What is Cognee?

Cognee is a graph-based memory engine for AI agents that captures information, organizes it as interconnected knowledge, and enables agents to recall context across conversations. It runs locally, on managed cloud infrastructure, or on-premises with full control over data.

Independently observed

Cognee pricing

Plans, per-tier features and add-ons, dated and linked to live pricing. Pricing changes often; always verify at source before you rely on it.

Pricing as of verify at live pricing ↗Independently observed
from $5/workspace/moHybridFree tier

Free tier (1M tokens/mo). Standard from $5/workspace/mo. Enterprise custom pricing.

Free

Free
Free

Self-hosted open-source deployment

users
unlimited
api_calls
unlimited
workspaces
1
tokens_monthly
1M
  • Graph-based memory
  • SDK and MCP access
  • Agentic integrations (Claude Code, Cursor, Codex)
  • Unlimited users and API calls

Standard

$5 per workspace/month
monthly

Managed cloud with data integrations

workspaces
unlimited
  • Unlimited workspaces
  • Data integrations: Slack, Notion, Google Drive
  • In-app support

Premium

Custom pricing with enterprise support
monthly

Enterprise deployment with dedicated support

  • Dedicated Slack channel
  • Dedicated support engineer
  • Bring-your-own cloud support
  • Support SLA

What Cognee does

The capabilities that matter for agent memory, normalised so it lines up with every alternative. “-” means we haven't confirmed it, not that it's missing.

Capabilities
Architecture model
Open source self hosted
Graph based entity relationship extraction
Automatic context window rolling summarization
Semantic vs episodic event chronological memory
Cross session and multi thread user persistence
Memory conflict resolution and fact updating
Multi tenant user and agent isolation
Native langchain and llamaindex memory classes
Time decay and recency weighted retrieval
Zero data retention and pii redaction layer
SOC2 type ii
Mit or apache permissive oss license
Pricing model
-
Independently observed

Platform & deployment

Independently observed
Platforms
  • CLI
  • Web
Deployment
  • Cloud / SaaS
  • On-premise
  • Air-gapped
  • Self-hosted

Integrations (27)

Independently observed
  • Claude Code
  • Cursor
  • Codex
  • LangGraph
  • OpenClaw
  • CrewAI
  • n8n
  • Slack
  • Notion
  • Google Drive
  • Amazon Neptune
  • Weaviate
  • Qdrant
  • Pinecone
  • Redis
  • Azure AI Search
  • Milvus
  • OpenSearch
  • Valkey
  • Memgraph
  • NetworkX
  • Google Cloud Spanner
  • TuringDB
  • pgGraph
  • DuckDB
  • FalkorDB
  • ScrapeGraphAI

Cognee alternatives

Other agent memory we track, ranked by the same independent score.

All Cognee alternatives, ranked →

Compare Cognee

Side by side against other agent memory, attribute by attribute, with a source on every value.

Independent · unbought · dated

The Vioscale score: one lens on the evidence

Not user reviews and not a paid placement: a confidence-weighted blend of the independent signals below (adoption, activity, security posture, and more), which you can sort and re-weight yourself. Vendors can correct their listing but can never move their rank, and stars are weighted low as a vanity metric. It is one way to read the evidence for Cognee, not the verdict.

Balanced composite 83 / 100
low · 37%
Signal contributions to the composite score
SignalScoreWeightContributionEvidence
Pricing transparency1000.088.4
Capabilities920.054.5
Price level800.054.2
Integrations420.041.7
Reliability00.070.0-
Security posture50.070.0-

Computed . Re-weight it by intent, or see the full method.

All data & sourcesshow ↓

Every value we hold, with its source, retrieval date, and confidence. This is the evidence behind the score: don't trust it, verify it.

Features

AttributeValueEvidence
CapabilitiesSoc2 type ii: No · Architecture model: open_source_self_hosted · Mit or apache permissive oss license: Yes · Multi tenant user and agent isolation: Yes · Time decay and recency weighted retrieval: Yes · Graph based entity relationship extraction: Yesmediumsource · 2026-08-21 · 60%

Integrations

AttributeValueEvidence
Count27mediumsource · 2026-08-21 · 60%

Market

AttributeValueEvidence
AvailabilityHqCountry: DE · PrimaryMarkets: … · AvailabilityScope: global · AvailableCountries: … · NotAvailableCountries: …highsource · 2026-08-21 · 75%

Pricing

AttributeValueEvidence
Free tierYesmediumsource · 2026-08-21 · 60%
Modelfreemiummediumsource · 2026-08-21 · 60%
TransparentYesmediumsource · 2026-08-21 · 60%
Price levellowmediumsource · 2026-08-21 · 60%
Starting priceAmount: 5 · Currency: USDmediumsource · 2026-08-21 · 60%

Security

AttributeValueEvidence
GdprYeshighsource · 2026-08-21 · 75%