The open-source alternatives to Glean are Onyx, AnythingLLM, Open WebUI and Khoj. Onyx, formerly Danswer, is the closest: more than 50 connectors, agentic retrieval over a hybrid index, deep research and custom agents, with an MIT Community Edition and enterprise sign-on sold separately. AnythingLLM and Open WebUI cover document chat with citations and roles for smaller teams, and Khoj answers from your notes and files as well as the web. A company switches to keep its documents and its model in-house.
3 of 4 carry an OSI licence. 4 of 4 run on your own machines. Figures read from GitHub on September 17, 2026.
One project per need, with the reason from its own README and the one thing to know before you choose it.
Formerly Danswer: 50+ connectors, agentic retrieval over a hybrid index, deep research and custom agents, with an MIT Community Edition and any local or hosted model.
Know firstSingle sign-on, SCIM and role-based access are paid Enterprise features.
Document search across 9 vector databases with hybrid retrieval, web search, and roles, LDAP, OAuth and SCIM in the free edition.
Know firstSource available: its branding may not be removed above 50 users without a paid licence.
MIT, chatting with your documents with source citations over 39 model providers, as a desktop app or a multi-user Docker server.
Know firstMulti-user support is Docker only, and telemetry is on until you turn it off.
AGPL-3.0, answering from your documents and the internet with any model, with agents, scheduled research and clients for Obsidian and WhatsApp.
Know firstIts README does not describe citations or name its search engine.
The facts that decide most choices. Each project's page carries the full panel.
| Project | Best for | Openness | Licence | Runs as | Own machines | Stars |
|---|---|---|---|---|---|---|
| Onyx | Company-wide search with agents | OSI | See LICENSE | Install script, Docker, Kubernetes (Helm, Terraform), Hosted cloud | Yes | 32,149 |
| Open WebUI | Documents plus roles and single sign-on | Source available | See LICENSE | pip, Docker, Kubernetes | Yes | 152,426 |
| AnythingLLM | Document chat with citations | OSI | MIT | Desktop app, Docker, Bare metal | Yes | 66,155 |
| Khoj | Notes, files and the web together | OSI | AGPL-3.0 | Self-hosted server, Hosted app, Browser, Obsidian, Emacs, desktop, phone and WhatsApp clients | Yes | 37,399 |
Licence and star figures read from GitHub on September 17, 2026. Stars are shown as a dated fact and were not used to rank this page.
An AI platform, formerly Danswer, that searches your company's knowledge through 50+ connectors with agentic retrieval, deep research, agents and actions, using any hosted or self-hosted model, with an MIT Community Edition.
Formerly Danswer: 50+ connectors, agentic retrieval over a hybrid index, deep research and custom agents, with an MIT Community Edition and any local or hosted model.
Know firstSingle sign-on, SCIM and role-based access are paid Enterprise features.
A self-hosted chat platform for Ollama and OpenAI-compatible models with document search over 9 vector databases, web search, image generation, roles and single sign-on, under a BSD-3 licence with an added branding clause.
Document search across 9 vector databases with hybrid retrieval, web search, and roles, LDAP, OAuth and SCIM in the free edition.
Know firstSource available: its branding may not be removed above 50 users without a paid licence.
An MIT all-in-one AI app from Mintplex Labs for chatting with your documents with citations, running agents with MCP, and using any of 39 model providers including local llama.cpp and Ollama, as a desktop app or a multi-user Docker server.
MIT, chatting with your documents with source citations over 39 model providers, as a desktop app or a multi-user Docker server.
Know firstMulti-user support is Docker only, and telemetry is on until you turn it off.
An AGPL-3.0 personal AI that answers from the web and your own documents with any local or hosted model, builds custom agents, sends scheduled research by email, and runs self-hosted or in its maintainers' cloud.
AGPL-3.0, answering from your documents and the internet with any model, with agents, scheduled research and clients for Obsidian and WhatsApp.
Know firstIts README does not describe citations or name its search engine.
The questions that settle it, in the order they usually come up.
Onyx lists 50+ connectors. The others work from documents you upload or point them at.
Open WebUI has it free; Onyx sells it in its Enterprise Edition.
All four support local models, so documents and model can stay in your network.
Glean, in the vendor's own words, is "the Enterprise AI platform connected to your enterprise's data. Find, create, and automate anything", by its own description. glean.com. The cases below are where that still wins.
Stars, licence, last commit and last release are read from the GitHub API by a script every week and carry the date they were read. Nobody types them.
What each project does is written from its own README and LICENSE, read in full on the date shown, and phrased as the project's claim. The sources are listed at the end.
That a project is faster, better or cheaper than Glean. A price for anyone. A user count. A roadmap. If a fact is not on this page, it was not in the source.
Onyx, formerly Danswer, with 50+ connectors, agentic retrieval, deep research and custom agents, free under MIT in its Community Edition. AnythingLLM and Open WebUI handle document chat with citations for smaller deployments, and Khoj searches your notes and files alongside the web.
Not the way a hosted service mirrors each source's permissions. Open WebUI has roles, groups, LDAP and single sign-on in its free edition; Onyx puts single sign-on, SCIM and role-based access in its paid Enterprise Edition; AnythingLLM has multi-user permissions in its Docker version.
Yes. Onyx runs self-hosted with models through Ollama, LiteLLM or vLLM; AnythingLLM and Open WebUI both work with local models. The documents, the index and the model can all sit on your own servers, which is the usual reason to move off a hosted service.
Onyx lists more than 50 indexing connectors out of the box or through MCP. The others work from documents you upload or point them at: AnythingLLM supports many file types with citations, Open WebUI searches across 9 vector databases, and Khoj reads PDF, Markdown, Word and Notion files.
The core of each is. Onyx's Community Edition is MIT, AnythingLLM is MIT, Khoj is AGPL-3.0, and Open WebUI is BSD-3 with a branding clause that applies above 50 users. Onyx's enterprise features and Khoj's enterprise edition are paid.
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.