xerj vs llm-app
Two of the top projects in RAG and search, side by side: who leads on each score part and who each one suits, then setup, license, activity and what each review found.
xerj
Elasticsearch-compatible search server that auto-indexes folders for coding agents
llm-app
Pathway RAG pipeline templates that re-index live data sources
Verdict
xerj ranks higher, 61 to 52. Of the 5 score parts we measure for both, xerj leads on 3, llm-app on 1 and 1 is even.
- Adoptionllm-app leads by 40 points, 95 to 55: a wide gap.
- FreshnessEven at 100.
- Maintenancexerj leads by 50 points, 99 to 49: a wide gap.
- Easy to runxerj leads by 17 points, 17 to 0: a wide gap.
- Agent-readyxerj leads by 30 points, 30 to 0: a wide gap.
Pick xerj if you are
Developers giving coding agents local code and document search or memory
- Single static binary, no JVM; prebuilt for Linux, macOS and Windows on x86-64 and arm64
- Works with existing Elasticsearch clients pointed at localhost:9200; 1366/1369 ES-YAML conformance tests
Pick llm-app if you are
Teams running RAG over documents that change constantly
- No separate vector DB, cache or API framework; indexing is in-process (usearch, Tantivy)
- Connectors for file system, Google Drive, SharePoint, S3, Kafka and Postgres with live sync
| What we compare | xerj | llm-app |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 55, popular | 95, widely used |
| Freshness | 100, active | 100, active |
| Maintenance | 99, healthy | 49, patchy |
| Easy to run | 17, hard | 0, hard |
| Agent-ready | 30, minimal | 0, none |
| Facts from GitHub and the README | ||
| Stars | 5.6k | 58.8k |
| License | Apache-2.0 (permissive) | MIT (permissive) |
| Last commit | Oct 2026 | Oct 2026 |
| Last release | Oct 2026 | None published |
| Language | Rust | Python |
| Docker | No | No |
| GPU | Not needed | Not needed |
| arm64 or Apple Silicon | Mentioned | Not stated |
xerj
XERJ is a single Rust binary that speaks the Elasticsearch API. `xerj autoindex <folder>` infers datasets from code, docs, logs, PDFs, SQLite and CSV, and creates one typed index per dataset, with tree-sitter symbols and line numbers for code. It supports BM25, kNN and hybrid search, a `/_memory` agent memory API, and a `xerj mcp` stdio server exposing 13 tools. Its main pitch is cutting agent token use by retrieving passages instead of reading files.
Who it is for: Developers giving coding agents local code and document search or memory
Strengths
- Single static binary, no JVM; prebuilt for Linux, macOS and Windows on x86-64 and arm64
- Works with existing Elasticsearch clients pointed at localhost:9200; 1366/1369 ES-YAML conformance tests
- Autoindex needs no schema; handles code, PDF, DOCX, XLSX, SQLite, mail and logs
- Default lexical embedder runs offline; air-gapped deployment recipe is documented
Weaknesses
- Token savings come from the project's own 16-run study; neutral to negative on popular libraries
- Default embedder is lexical feature-hashing; neural mode is opt-in and downloads about 90 MB
- `--insecure` quick start disables TLS and auth; every caller becomes a superuser
- Optional Jev rerank sends result text to TypeSafe's servers
- Rust, lexical embedder (default, offline), neural embedder (opt-in, ~90 MB download), Jev reranker (optional, external service)
llm-app
Eight Dockerized Python pipelines on the Pathway framework: question-answering RAG, a live document indexer, multimodal RAG with GPT-4o, unstructured-to-SQL, adaptive RAG, a private Mistral plus Ollama variant, slide search and video RAG. Each watches a source (file system, Google Drive, SharePoint, S3, Kafka, Postgres), keeps an in-memory vector and full-text index current and serves an HTTP API. For teams whose documents change constantly.
Who it is for: Teams running RAG over documents that change constantly
Strengths
- No separate vector DB, cache or API framework; indexing is in-process (usearch, Tantivy)
- Connectors for file system, Google Drive, SharePoint, S3, Kafka and Postgres with live sync
- Private variant runs fully local with Mistral and Ollama
- Docker images and tests included
Weaknesses
- Pathway is the real dependency; its Rust engine is opaque to most Python teams
- Pipelines are backends; the UI is an optional Streamlit demo
- Root README has no setup; each template README is required reading
- Index lives in memory; sizing for millions of pages is on you
- Python, pathway, openai, mistral, ollama
- Needs docker, openai-api-key, data-source-credentials
More in RAG and search
Building with these?
We help teams pick an open-source AI stack and run it in production. Client work never changes a rank.