leviathan vs xerj
Single Rust binary that turns JSONL, CSV, SQLite or any DB export into a ranked full-text index, giving agents short cited answer cards (~450 tokens) at any dataset size. — versus — One Rust binary for agent search: BM25, vector and hybrid search over code, docs, logs and PDFs, an Elasticsearch-compatible API, MCP, and Query-1, which answers by running programs over facts.
Both are single Rust binaries that index records so agents get cited, token-cheap answers. Leviathan is a lean full-text CLI; XERJ a full engine with vector search, a server and an Elasticsearch-compatible API.
| leviathan | xerj | |
|---|---|---|
| Stars | 879 | 19k |
| Forks | 53 | 1.4k |
| Language | Rust | Rust |
| License | Apache-2.0 | Apache-2.0 |
| Last activity | 4 days ago | today |
| Topics | memory, data | rag, storage |
| Curated connections | 3 | 3 |
leviathan — the curator's take
Use it when an agent has to answer questions over large record dumps, like tickets, logs or CRM exports, and grep is flooding the context. The benchmark claims 245x fewer tokens than the best grep strategy. Skip it if you need semantic or vector recall: it is lexical full-text search. It is not a memory system that learns from conversations; you re-index when the data changes.
xerj — the curator's take
Use it when you want a single self-hosted search engine for agents across code, docs and data instead of Elasticsearch plus a vector DB, or when wrong answers are costly and you want UNKNOWN or CONFLICT instead of guesses. Skip it if you just need grep-replacement for one repo, since it is a whole engine to operate. The headline benchmark numbers come from the project itself.