Skip to content
InfratoApp
InfratoApp / Resources / Open-source alternatives / deepset

deepset

deepset-ai/haystack

An open-source alternative to AWS Kendra and QnA Maker.

Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Built for scalable agents, RAG, multimodal applications, semantic search, and conversational systems. (from the repository)

Stars
26,762
Licence
Apache-2.0
Last push
2026-10-17
Written in
Python
Source on GitHub ↗Project website ↗

Is it alive?

Four readings from the repository, each with the rule behind it. No blended score.
Activity
Active
Last push 10 days ago
Active: pushed within 90 days. Slowing: within a year. Quiet: longer.
Releases
Recent
v3.4.0, 12 days ago
Recent: a tagged release within six months. Some projects ship from the main branch and never tag.
Issues
98% closed
89 open · 4,508 closed
Share of all issues ever filed that are now closed. Under 20 issues is too few to read.
Momentum
+90 in 9 days
26,762 stars
Change in stars across our daily readings, over up to 30 days. Needs a week of readings.

What a business may do with it

PermissiveApache-2.0

Use, change and sell it, including inside a closed product. Keep the copyright notice.

This describes the licence family GitHub identified, not legal advice about this project. The repository’s own LICENSE file is what binds you.

Run it yourself

deepset’s README does not show a docker run, docker compose up or helm install in a code block, so this page will not guess one. It may be a desktop app, a library, or documented on its website.

deepset’s own install instructions ↗

Where this page comes from

  • Stars, licence, activity, issues, releases and the self-host command are read from the repository. 2026-10-18
  • The pairing with AWS Kendra and QnA Maker and the category come from opensourcealternative.to, a community directory of open-source alternatives. We re-read it weekly.
  • Star momentum comes from our own daily readings of the repository. How we collect data