Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
Supermemory is a “context cloud” for AI Agents. Rather than being just a vector database, it combines long-term and short-term memory, RAG, user profiles, connectors, content extraction, and knowledge graphs into a single API. Developers can send in text, files, chat logs, and more; the system extracts the content and builds a semantic understanding graph around users, documents, projects, or organizations, then returns relevant context at query time for use with any model.
Its Memory API is designed around the idea that facts evolve over time: it can merge, update, handle contradictions, forget information, and generate user profiles. SuperRAG offers hybrid search, reranking, structured context, and an officially claimed latency of under 300ms. Extractors support rich content such as PDFs, web pages, images, audio, and video. Connectors cover Slack, Notion, Drive, Gmail, GitHub, S3, and more. It also provides TypeScript/Python SDKs, MCP, plugins, filesystem mounting, and a Chrome Extension, making it suitable for integration into Agent workflows such as Claude, Cursor, Codex, and OpenCode AI.
Pricing is fairly clear: the Free plan includes $5 in monthly credits; Pro is $19/month, Max is $100/month, Scale is $399/month, and Enterprise is custom. Under the hood, billing is based on SM tokens, queries, and operations, with deduplication for repeated content, so duplicate uploads are not charged again. Memory starts from $0.005/1K SM tokens, and SuperRAG starts from $0.001/1K. Startups and research projects can apply for $1000 in credits.
The main advantage is its complete product stack, which reduces the cost and complexity of teams building their own combination of RAG, memory systems, extraction services, and sync pipelines. It supports any model and multiple deployment options, making it friendly for production-grade Agents. On the security side, it offers SOC 2, GDPR, self-hosting, VPC, and air-gapped options. Limitations include the fact that Chinese language support is not clearly stated in official materials, and its benchmarks and low-latency claims still need to be validated on real-world data. Advanced connectors, self-hosting, and compliance features are mostly available on Scale/Enterprise plans, while the free quota is limited for individuals or small teams.
Supermemory is suitable for development teams building personalized AI assistants, enterprise knowledge bases, real-time knowledge synchronization, internal document search, and user profiling systems. If you only need lightweight Q&A or static document retrieval, a traditional vector database or simple RAG setup may have a lower barrier to entry. Access from mainland China, supported payment methods, and network stability are not specified in the available materials, so they should be considered unknown. Alternatives to compare include Mem0, Zep, or a self-built vector database/RAG solution.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on supermemory.ai official site.
supermemory.ai is an Unknown AI Apps provider. TG4G tracks its product information, an overall rating of 8.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach supermemory.ai directly.