Cognee alternatives
Graph memory when documents→ontology matter. Zep for temporal bipartite graphs; Mem0 for simplest hosted…
Overall 8.09 (curated_estimate, high) · score asOf 2026-09-14 · last reviewed 2026-09-14. This is not an Elo and not a live price board. Provider · JSON
Jobs: rag-memory, hosted-agent-memory, temporal-graph-memory, embedded-vector, cross-session-user-memory, knowledge-graph-rag
| Provider | Why listed | Overall | Δ vs subject | asOf |
|---|---|---|---|---|
| Zep Pick when agents need temporal/entity graph memory, not just vector snippets. Mem0 for simpler recall. | same_jtbd_runner, same_jtbd_pick · hosted-agent-memory, temporal-graph-memory, cross-session-user-memory | 7.44high | -0.65 | 2026-09-14 |
| Mem0 Default hosted agent memory bolt-on. Model retrieval volume carefully; self-host OSS if caps hurt. | same_jtbd_pick, same_jtbd_runner · hosted-agent-memory, temporal-graph-memory, cross-session-user-memory | 7.59high | -0.5 | 2026-09-14 |
| Supabase (pgvector) Depot default for most agent apps under ~25–50M vectors. Don't add Pinecone until Postgres actually hurts. | same_jtbd_pick, same_jtbd_runner · rag-memory, hosted-agent-memory | 8.64high | +0.55 | 2026-09-14 |
| Memobase Depot pick for portable/user-profile memory (MCP across Claude/ChatGPT/agents). Mem0 remains general… | same_jtbd_runner, same_jtbd_pick · hosted-agent-memory, cross-session-user-memory | 7.94high | -0.15 | 2026-09-14 |
| LanceDB Great embedded/object-store vector layer. Pair with Mem0/Zep only if you need higher-level memory APIs. | same_jtbd_pick · embedded-vector | 7.78medium | -0.31 | 2026-09-14 |
| Turbopuffer When your vector bill scares you and much of the corpus is cold, Turbopuffer is the interesting answer. | same_jtbd_runner · rag-memory, embedded-vector | 8.21high | +0.12 | 2026-09-14 |
| Weaviate Best when hybrid search and modularity matter. Cloud for convenience; self-host when lock-in anxiety spikes. | same_jtbd_runner · temporal-graph-memory, knowledge-graph-rag | 8.2high | +0.11 | 2026-09-14 |
| Supermemory Managed context engine when you want memory without running a vector DB. Mem0 for simpler fact API; Cognee… | same_jtbd_runner · hosted-agent-memory, cross-session-user-memory | 7.88high | -0.21 | 2026-09-14 |
| Letta Cloud Choose when the agent IS a long-lived stateful runtime. For bolt-on memory use Mem0/Zep. | same_jtbd_runner · hosted-agent-memory, temporal-graph-memory | 7.4medium | -0.69 | 2026-09-14 |
| Qdrant Depot pick for serious vector workloads with filters. Self-host or Cloud — either way you can leave. | same_jtbd_runner · rag-memory | 8.59high | +0.5 | 2026-10-19 |
| Upstash Redis Default serverless-edge-redis. Use classic Redis Cloud when you need always-on fat memory / Redis modules… | same_jtbd_runner · rag-memory | 8.49high | +0.4 | 2026-09-14 |
| Upstash Vector Prototype/edge RAG peer when you already on Upstash. Prefer Qdrant/Pinecone for serious filtered production… | same_jtbd_runner · rag-memory | 8.15high | +0.06 | 2026-09-14 |
Prices are not ranked on this page. For a job, use the sourced quote (separate input and output rows, Depot Pick unchanged): rag-memory, hosted-agent-memory, temporal-graph-memory, embedded-vector.
Alternatives are other catalog providers that share a JTBD as pick or runner-up. If fewer than three exist, same-category active providers fill the list and are labeled same_category. Overall deltas are catalog scores. kind=curated_estimate is not a lab measurement. Pricing rows are sourced or explicitly null. Do not parse pricingNotes. A single normalized USD figure is set only for a provider with exactly one token row.