Stay free if you only need 1m tokens/month and 10k search queries/month. Upgrade if you need 80m tokens/month and 20m search queries/month. Most solo builders can start free.
Why it matters: Scale tier jumps sharply from $19/month Pro to $399/month, leaving a large gap for mid-sized teams
Available from: Pro
Why it matters: Gmail, S3, and Web Crawler connectors are gated to the $399 Scale tier and above
Available from: Pro
Why it matters: Overage charges ($0.01 per 1,000 tokens, $0.10 per 1,000 queries) can add up for unpredictable workloads
Available from: Pro
Why it matters: As a newer memory-layer category, best practices and community tutorials are still maturing compared to established vector DBs
Available from: Pro
Supermemory is not another vector database â it is a custom-built engine that combines a Vector Graph Engine with a User Understanding Model. Unlike pure vector stores that only compute similarity scores, Supermemory maps ontology-aware edges that represent real relationships between memories, and builds behavioral profiles of users from their interactions. This means agents can retrieve not just semantically similar chunks but contextually connected knowledge, including user intent and preferences. It also bundles connectors, extractors, and retrieval in a single API so teams don't have to stitch together five services.
Supermemory has four tiers: Free ($0 with 1M tokens/month and 10K queries/month), Pro ($19/month with 3M tokens and 100K queries plus all plugins), Scale ($399/month with 80M tokens, 20M queries, and Gmail/S3/Web Crawler connectors), and Enterprise (custom pricing with unlimited usage, forward-deployed engineer, SSO, and custom integrations). All plans include unlimited storage, unlimited users, and free multi-modal extraction. Overages on Pro and Scale are charged at $0.01 per 1,000 tokens and $0.10 per 1,000 queries. Qualifying startups can apply for $1,000 in credits and 6 months of dedicated support.
Yes. The Enterprise plan supports self-hosting inside your own VPC and cloud environment, giving you full control over infrastructure and data residency. Supermemory is also certified to SOC 2, HIPAA, and GDPR standards. The company explicitly states it does not train models on customer data and that you can export your data at any time. This makes it viable for regulated industries like healthcare, finance, and legal tech that cannot send data to third-party SaaS.
Supermemory ships with SDKs in TypeScript, Python, and a REST API, plus native integrations with Claude Code, OpenClaw, OpenCode, Vercel AI SDK, LangChain, LangGraph, CrewAI, OpenAI SDK, Mastra, Zapier, n8n, and Pipecat. There are also consumer plugins including a Chrome extension and desktop apps for saving links, chats, PDFs, images, and videos. This range of 14+ integrations means teams can adopt Supermemory without rewriting their existing agent stack â three lines of code are typically enough to add it to an existing LangChain or CrewAI project.
Supermemory is best suited for three audiences: AI developers building agents that need long-term memory across sessions; startups and scale-ups that need production-grade retrieval with sub-300ms latency without building it in-house; and enterprises requiring self-hosted, compliant memory infrastructure for regulated workloads. Individual power users (10,000+ of them) also use the Personal Supermemory app to unify memory across Claude, Cursor, ChatGPT, and other assistants. Teams that only need basic RAG over a small document set may find it more than they need, while those juggling multiple memory tools will benefit from the consolidated API.
Start with the free plan â upgrade when you need more.
Get Started Free âStill not sure? Read our full verdict â
Last verified March 2026