Simple, transparent pricing

Open source is free. Managed Cloud scales with you. Enterprise gets dedicated support.

Preview pricing

Brain Cloud is in early access. The prices, SLA, and performance figures below are targets — final pricing is set at general availability.

Open Source

Free

Self-hosted, Apache 2.0 licensed. Perfect for local development and self-hosted deployments.

Memory SDK for Python, TypeScript, Rust
Full source code on GitHub
5 retrievers included
Hybrid search + retrieval fusion
Community Discord support

Managed Cloud

Pay-as-you-go

API-first memory service with auto-scaling, multi-region availability, and compliance features.

Everything in Open Source +
Global API, p99 < 200ms (target)
Automatic scaling & backups
SOC 2 Type II in progress
GDPR-ready
Built-in audit logs
Email support

Enterprise

Custom

Dedicated infrastructure, SLA, SSO, and priority support for large-scale deployments.

Everything in Cloud +
Dedicated infrastructure option
99.9% uptime SLA (planned)
SAML/SSO integration
Priority support & SLAs
Custom integrations

How Brain Compares

Estimated monthly costs for typical agent memory scenarios. Estimates based on public pricing; actual costs may vary. Check each provider's current pricing page for exact rates. Brain Cloud figures are preview/target pricing, not final GA rates.

Scenario 1: 100K vectors, 1,000 queries/month

Early-stage agent with moderate scale. Typical cost per provider:

Provider	Storage	Retrieval	Total/month
Brain Cloud	$5	$1	$6
Mem0	—	$20	$20
Pinecone	$8	$0	$8

Assumptions: Brain (preview): $0.05/1K vectors stored + $0.001/query. Mem0: $0.002/API call, 10K calls. Pinecone: Standard index with 100K vectors.

Scenario 2: 1M vectors, 10K queries/month

Production agent with significant scale. Typical cost per provider:

Provider	Storage	Retrieval	Total/month
Brain Cloud	$50	$10	$60
Mem0	—	$200	$200
Pinecone	$73	$0	$73

Assumptions: Brain (preview): $0.05/1K vectors + $0.001/query. Mem0: $0.002/call, 100K calls. Pinecone: Standard index with 1M vectors, 10K queries.

What's included

Brain Cloud:Managed vector storage, hybrid retrieval (BM25 + semantic), SOC 2 Type II (in progress), GDPR-ready, multi-region replication, p99 < 200ms (target)

Mem0: Memory structuring (entity extraction, relationship mapping), high-level memory abstractions, SaaS simplicity, API-first design

Pinecone: Fully managed vector database, serverless infrastructure, querying included in index cost, global availability

Frequently asked

Can I self-host the Cloud version?

The Cloud API is managed infrastructure only. For self-hosting, use the open source version.

What's included in the free tier?

Full SDKs (Python, TypeScript, Rust), retrieval fusion, 5 built-in retrievers, community support on Discord.

Does Cloud have a free trial?

Early access is available — join the waitlist from your dashboard after signing up.

Can I migrate from open source to Cloud?

Yes, the API is fully compatible. Point your SDK to the Cloud endpoint and run migrations with zero code changes.

Questions about licensing?

See our full license on GitHub or contact us for Enterprise and compliance details.