Skip to content
Get packs

Quivr

Open-source engine that ingests content streams and makes them searchable and monitorable, in evaluation stage

GitHub Download v2.0.2
Stars
39.6k
Forks
3.7k
Licence
MIT
Language
Go
Last update
4 days ago
Type
Framework

What it is

Quivr V2 is a Go-based engine that takes in content continuously, makes it searchable quickly, adds enrichments such as embeddings in the background, and lets you follow a topic over time with alerts. It stores canonical data in PostgreSQL and S3 and uses Weaviate for search, with plugins meant to handle formats, AI models and business rules. The README says it is at an evaluation stage and not yet ready for production.

  • Create a Corpus, add an article, search it, and receive an alert, as described in the Quickstart
  • Run a local stack with make dev and inspect the API
  • Use the optional browser UI over the same API via make demo
  • Build plugins for custom formats, models or delivery channels

Great for

  • Developers building a search or monitoring pipeline over a stream of documents
  • Teams who want to evaluate a self-hosted, plugin-based ingestion and search engine
  • People comfortable running Docker Compose and reading API documentation
  • Engineers who want durable, idempotent writes and rebuildable search indexes

Not ideal for

  • Anyone needing a production-ready system, since the README says the API is v0 and may change and should not yet run in production
  • Non-technical users looking for a ready-made chat or end-user app, as the README describes an engine and API
  • Users on Windows or macOS Intel, since the quickstart lists Linux x86_64 or macOS on Apple Silicon only
Step-by-step guide: how to use QuivrInstall, first run, examples, tips and FAQs
LangChain vs Quivr

How to get it

  1. 1

    Clone and build from source

    You'll need: Go 1.27.2 · Docker with Compose v2 · Python 3 with venv · Node.js 22+ · jq

    Clone the repository

    git clone https://github.com/The-Vibe-Company/quivr

    Start dependencies, run migrations and launch API and worker

    make dev

    Run checks without Docker

    make check
  2. 2

    Run it

    • Run make dev and note the API address and generated config.json path printed in the output
    • Run eval "$(make -s env)" to export the address, key and webhook destination
    • Follow the Quickstart to create a Corpus, add an article, search it and get an alert
    • Optionally run make demo and open http://127.0.0.1:5183 for the browser UI
  3. 3

    Get the most out of it

    • Use make verify to run the end-to-end journeys on an isolated stack and get a report
    • Add plugins for formats, models or business rules rather than changing the core
    • Configure weaviate_api_key and HTTPS when using an external vector index
    • Use scoped API keys per Organization, action and Corpus to control access
    • Review the OpenAPI contract in contracts/http/v0 when building integrations

Downloads

Source code (.zip)main branch · straight from GitHubRelease v2.0.28 files incl. quivr-compose-2.0.2.tar.gzOfficial website & docswww.thevibecompany.co/en/portfolio/quivr

Strengths

  • Writes are acknowledged only after commit, and the README says it avoids silent failures
  • Search indexes can be rebuilt from PostgreSQL and S3 canonical data
  • Released under the MIT license with an OpenAPI contract in the repository

Trade-offs

  • The README marks it as evaluation stage with an unstable v0 API
  • Setup requires several dependencies (Go, Docker, Python, Node.js, jq) and a first-run download of about 1 GB
  • The latest release has no binaries, so installation is from source or images

How it compares

ProjectStarsTypeLevelUpdatedBest for
quivr (this)39.6kFrameworkadvanced4 days agoDevelopers building a search or monitoring pipeline over a stream of documents
postgrest27.7kAppintermediate6 days agoDevelopers who already have a PostgreSQL database and want an HTTP API on top of it quickly
one-api37.3kAppintermediate10 months agoDevelopers who want one OpenAI-compatible endpoint for several model providers
repomix29.1kCLIbeginner4 days agoDevelopers who want to ask an AI assistant questions about a whole project
botpress15kDev toolintermediate6 days agoDevelopers who want to write integrations and bots as TypeScript code using the Botpress CLI and SDK
full-stack-fastapi-template46.1kFrameworkintermediate4 days agoDevelopers starting a new full-stack web app with a Python backend and React frontend

Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.