Ollama
Run open LLMs locally with a single command.

Open-source AI chat and enterprise search over your company's knowledge.
Onyx is one of the strongest open-source options in enterprise search, and its MIT license, 40+ connectors, and model-agnostic design make it genuinely appealing for teams that want data control without vendor lock-in. The honest tradeoff is that self-hosting requires engineering effort to deploy and maintain; teams wanting turnkey simplicity may prefer the paid cloud. Verify current cloud pricing with Onyx.
Onyx (formerly Danswer) is an open-source, MIT-licensed enterprise search and AI assistant that connects to 40+ workplace apps and answers questions in natural language over internal knowledge. It is model-agnostic, works with any LLM, and can be self-hosted for free or run as a managed cloud service priced per user. It appeals to security-conscious, engineering-capable teams that want data control and no vendor lock-in.
Onyx, previously known as Danswer, is an open-source AI chat and enterprise search platform that lets organizations ask questions in natural language across their internal documents, apps, and workflows. It ships with 40+ connectors (including tools like Microsoft Teams, Slack, Confluence, and Google Drive) and a model-agnostic architecture that works with essentially any LLM, so teams can pick their provider and keep control of their data. Because it is MIT-licensed and open-source, Onyx can be self-hosted at no software cost, giving organizations data residency, granular access controls, and the ability to customize connectors and behavior. This makes it especially attractive to security-conscious teams that want an alternative to closed enterprise-search SaaS. For teams that prefer not to run infrastructure, Onyx also offers a managed cloud option priced per user. Onyx emerged from Y Combinator (W24) and has grown a substantial open-source following on GitHub. It occupies the intersection of enterprise search, internal knowledge assistants, and open, self-hostable AI tooling, and is a strong fit for engineering-capable teams that value flexibility and control over turnkey simplicity.
Onyx (formerly Danswer) is an open-source, MIT-licensed enterprise search and AI assistant that connects to 40+ workplace apps and answers questions over internal knowledge. It is model-agnostic and can be self-hosted for free or run as managed cloud from around $20/user/month. It suits engineering-capable, security-conscious teams that want data control, but self-hosting requires operational effort. It came out of Y Combinator (W24).
Onyx, formerly Danswer, is a San Francisco-based company building an open-source AI search and assistant platform. Launched in 2023 and part of Y Combinator's W24 batch, it has grown a large open-source community on GitHub.
The project balances a free, self-hostable MIT-licensed core with a paid managed cloud, following a common open-core commercial model. Public sources place it in the small-team stage with an estimated enterprise value in the tens of millions; treat such figures as approximate.
Onyx offers natural-language search and AI chat over internal knowledge, 40+ connectors, model-agnostic LLM support, granular access controls, custom assistants, and data-residency options. It can be self-hosted via Docker/Kubernetes or consumed as managed cloud.
Because it is open-source, teams can customize connectors, control their data, and choose their own models, which is a core differentiator against closed enterprise-search SaaS.
Engineering-capable, security- and privacy-conscious organizations that want a customizable, self-hostable enterprise search and internal knowledge assistant with no vendor lock-in.
Employees across a company searching internal docs and apps via natural language.
Engineering leaders, IT, and heads of internal tooling choosing an enterprise search solution.
Security, compliance, and platform teams who value open-source and data residency.
A security-conscious, engineering-capable organization that wants customizable, self-hostable enterprise search over internal knowledge and prefers open-source flexibility to closed SaaS.
Onyx (Danswer) participated in Y Combinator's W24 batch. Beyond YC backing, no large disclosed venture rounds were confirmed at the time of writing; public trackers list it as an early-stage company. Verify funding details with primary sources.
The self-hosted, MIT-licensed version is free to use; you only pay for your own infrastructure and LLM API costs. Onyx also offers a paid managed cloud, reported to start around $20 per user per month.
Onyx was formerly known as Danswer. It rebranded to Onyx while continuing as an open-source enterprise search and AI assistant platform.
Onyx is model-agnostic and works with essentially any LLM provider, so teams can choose their own model and keep control of their data.
Onyx ships with 40+ connectors to common workplace tools such as Slack, Microsoft Teams, Confluence, Google Drive, Notion, and GitHub.
For self-hosting, yes, deploying and maintaining Onyx requires engineering capability. Teams that prefer no operations can use the managed Onyx Cloud instead.
Side-by-side pages for pricing, features, and best-fit use cases.
Run open LLMs locally with a single command.
Open-source AI-native vector database
High-performance open-source vector search engine
Open-source embedding database for AI apps