Onit – Source-available ChatGPT Desktop with local mode, Claude, Gemini
Hey Hackernews- it’s Tim Lenardo and I’m launching v1 of Onit today! Onit is ChatGPT Desktop, but with local mode and support for other model providers (Anthropic, GoogleAI, etc). It's also like Cursor Chat, but everywhere on your computer - not just in your IDE! Onit is open-source! You can download a pre-built version from our website: www.getonit.ai Or build directly from the source code: https://github.com/synth-inc/onit We built this because we believe: Universal Access: AI assistants should be accessible from anywhere on my computer, not just in the browser or in…
In plain words
Onit is a desktop application that brings AI chat capabilities to any part of a user's computer, supporting multiple model providers including ChatGPT, Claude, and Gemini. It offers a local mode for processing data privately without uploading to external servers, and works across applications rather than being confined to a browser or IDE. The software is open-source and available as a pre-built download or buildable from source code, giving users choice in both their AI provider and how they deploy the tool.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hey Hackernews- it’s Tim Lenardo and I’m launching v1 of Onit today! Onit is ChatGPT Desktop, but with local mode and support for other model providers (Anthropic, GoogleAI, etc). It's also like Cursor Chat, but everywhere on your computer - not just in your IDE! Onit is open-source! You can download a pre-built version from our website: www.getonit.ai Or build directly from the source code: https://github.com/synth-inc/onit We built this because we believe: Universal Access: AI assistants should be accessible from anywhere on my computer, not just in the browser or in specific apps Provider Freedom: Consumers should have the choice between providers (anthropic, openAI, etc.) not be locked into a single one (ChatGPT desktop only has OpenAI models) Local first: AI is more useful with access to your data. But that doesn't count for much if you have to upload personal files to an untrusted server. Onit will always provide options for local processing. No personal data leaves your computer without approval Customizability: Onit is your assistant. You should be able to configure it to your liking Extensibility: Onit should allow the community to build and share extensions, making it more useful for everyone. The features for V1 include: Local mode - chat with any model running locally on Ollama! No internet connection required Multi-provider support - Top models for OpenAI, Anthropic, xAI, and GoogleAI File upload - add images or files for context (bonus: Drag & drop works too!) History - revisit prior chats through the history view or with a simple up/down arrow shortcut Customizable Shortcut - you pick your hotkey to launch the chat window. (Command+zero by default) Anticipated questions: What data are you collecting? Onit V1 does not have a server. Local requests are handled locally, and remote requests are sent to model providers directly from the client. We collect crash reports through Firebase and a single "chat sent" event through PostHog analytics. We don't store your prompts or responses. How to does Onit support local mode? For use local mode, run Ollama. You can get Ollama here: https://ollama.com/ Onit gets a list of your local models through Ollama’s API. Which models do you support? For remote models, Onit V1 supports Anthropic, OpenAI, xAI and GoogleAI. Default models include (o1, o1-mini, GPT-4o, Claude3.5 Sonnet, Claude3.5 Haiku, Gemini 2.0, Grok 2, Grok 2 Vision). For local mode, Onit supports any models you can run locally on Ollama! What license is Onit under? We’re releasing V1 available on a Creative Commons Non-Commercial license. We believe the transparency of open-source is critical. We also want to make sure individuals can customize Onit to their needs (please submit PRs!). However, we don’t want people to sell the code as their own. Where is the monetization? We’re not monetizing V1. In the future we may add paid premium features. Local chat will- of course- always remain free. If you disagree with a monetized feature, you can always build from source! Why not Linux or Windows? Gotta start somewhere! If the reception is positive, we’ll work hard to add further support. Who are we? We are Synth, Inc, a small team of developers in San Francisco building at the frontier of AI progress. Other projects include Checkbin (www.checkbin.dev) and Alias (deprecated - www.alias.inc). We’d love to hear from you! Feel free to reach out at contact@getonit dot ai. Future roadmap includes: Autocontext - automatically pull context from computer, rather than having to repeatedly upload. Local-RAG - let users index and create context from their files without uploading anything. Local-typeahead - i.e. Cursor Tab but for everywhere Additional support - add Linux/Windows, Mistral/Deepseek etc etc. (maybe) Bundle Ollama to avoid double-download And lot’s more!
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, January 2025
the whole month →- IM
Hello! I'm Byran. I spent the past ~6 months engineering a laptop from scratch. It's fully open-source on GH at: https://github.com/Hello9999901/laptop
Dev tools · 2025 · byran.ee
- TITetris in a PDF▲1,289
I realized that the PDF engines of modern desktop browsers (PDFium and PDF.js) support JavaScript with enough I/O primitives to make a basic game like Tetris. It was a bit tricky to find a union of features that work in both engines, but in the end it turns out that showing/hiding annotation "fields" works well to make monochrome pixels, and keyboard input can be achieved by typing in a text input box. All in all it's quite janky but a nice reminder of how general purpose PDF scripting can be. The linked PDF is all ASCII so you can just open it in a text editor, or have a look at…
Life & fun · 2025 · th0mas.nl
Create lifelike, personalized AI avatars from text prompts
AI · 2025 · jogg.ai
