memvid vs minigraf
Single-file memory layer for agents: data, embeddings, index and metadata in one portable .mv2 — append-only Smart Frames, time-travel queries, sub-5ms recall, no server. Rust core, Node/Python SDKs. — versus — Embedded single-file graph database in Rust with Datalog queries and bi-temporal time travel — 'SQLite for connected data with history' — running native, WASM, mobile and in the browser.
Both give agents portable single-file memory; memvid is embedding/frame-based recall, minigraf a queryable temporal fact graph.
| memvid | minigraf | |
|---|---|---|
| Stars | 17k | 225 |
| Forks | 1.5k | 14 |
| Language | Rust | Rust |
| License | Apache-2.0 | Apache-2.0 |
| Last activity | 3 months ago | yesterday |
| Topics | memory, local | knowledge-graphs, memory, storage |
| Curated connections | 4 | 3 |
memvid — the curator's take
The portability bet is the real differentiator: one .mv2 capsule holds the whole memory — hand it to another agent, ship it offline, branch it, rewind it. The project began life literally encoding text into video frames; the codec DNA survived the Rust rewrite as compression smarts, and the maturity question survives with it. Benchmark claims (+35% SOTA on LoCoMo, '1,372× throughput') are self-run — the eval is open, so run it before repeating it. When NOT: team-shared concurrent memory — single file, append-only, no server means no multi-writer story; that's what the server-based layers are for.
minigraf — the curator's take
Use it when an agent or app needs local graph memory with full history (what was true, and when we learned it) in one file, no server. Skip it for vector/semantic recall (no embeddings), large multi-user workloads, or if Datalog is a non-starter; on v2.x, note the multi-valued-attribute bug and write each value in its own transact.