GodMode – open-source AI Chat Browser
Hi HN, we're Sean, Alec, and swyx, and we've been hacking on this side project that harmonizes all the major AI chat apps (Closed LLMs, Open Source Hosted LLMs, and Local/GGML ones too) and lets you invoke them in one keyboard shortcut. I originally made this as a smol menubar app that could quickly access ChatGPT from wherever I was on desktop. This necessitated writing an Electron app, rather than a Chrome extension. I also didn't have access to the GPT-4 API at the time, so I preferred to use GPT-4 via the full webapp. Then Bing came out strong, and Bard, and Claude, and Llama, and…
In plain words
GodMode is an open-source desktop application that consolidates access to multiple AI chat services into a single interface. It supports closed-source models like ChatGPT and Claude, open-source hosted options, and local models that run on-device. Users can invoke any of these AI assistants through a keyboard shortcut from anywhere on their computer, eliminating the need to switch between separate applications or websites. Built as an Electron app, it lets users quickly compare responses across different models.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hi HN, we're Sean, Alec, and swyx, and we've been hacking on this side project that harmonizes all the major AI chat apps (Closed LLMs, Open Source Hosted LLMs, and Local/GGML ones too) and lets you invoke them in one keyboard shortcut. I originally made this as a smol menubar app that could quickly access ChatGPT from wherever I was on desktop. This necessitated writing an Electron app, rather than a Chrome extension. I also didn't have access to the GPT-4 API at the time, so I preferred to use GPT-4 via the full webapp. Then Bing came out strong, and Bard, and Claude, and Llama, and Vicuna, and Falcon and on and on, and other people kept reporting amazing SOTA results on model after model and I felt hopelessly behind - either you just take benchmark results at face value, or you spend serious time copying and pasting chats just to get a feel for the practical differences between the various closed/open source LLMs. The real value came when I started just running every query past 3-5 chats at once, and started getting a sense for what each was good at. Bing/You.com is (obviously) good at search, but so is Perplexity. Claude 2 wins about 30% of my code queries vs GPT4. GPT4 and Bard and Bing handle file uploads differently. Llama2 is surprisingly good sometimes but also hopelessly bad at others. And so on. But the answer isn't to pick one daily driver. It's to make it easy to just run them all at once, and swap in/out whenever necessary. (I've taken to calling this Mixture of Mixture of Experts, akin to the GPT4 rumors geohot dropped on our podcast https://news.ycombinator.com/item?id=36413296) As the app grew in screen real estate (ultrawide monitor people! this is your calling!) and I started needing to copy/paste things more, I realized what we were really building was a full-fledged secondary browser that does a slightly different job than your regular browser. It's nominal first purpose is to just harmonize all chat inputs, but more broadly, I can pipe any arbitrary input to any open/closed LLM without any api key using your own account, and use it to enhance your browsing experience generally without me ever seeing your data. What could you do if you could add intelligence features for ~free to any part of your browsing experience? MS Edge has shown a bit of this potential but why should we be limited to just the Edge browser, or just the sidebar? (you can see the beginnings of our thinking in the alpha PromptCritic feature we just added). You can see our 1min launch video here: https://twitter.com/swyx/status/1692988634364871032 We've been hacking on this as a side project, and it's my first ever Electron app using a bunch of gnarly features. The UX has a lot of things to improve (see the open issues... help!) but just wanted to get this out there. Feedback welcome!
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 · 16d 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 · 26d ago · cactuscompute.com
Launched alongside, August 2023
the whole month →
Lottielab▲871Create and ship lottie animations to sites and apps faster
Dev tools · 2023 · lottielab.com
- LC
Outlines is a Python library that focuses on text generation with large language models. Brandon and I are not LLM experts and started the project a few months ago because we wanted to understand better how the generation process works. Our original background is probabilistic, relational and symbolic programming. Recently we came up with a fast way to generate text that matches a regex (https://blog.normalcomputing.ai/posts/2023-07-27-regex-guide...). The basic idea is simple: regular expressions have an equivalent Deterministic-Finite Automaton (DFA) representation. We…
AI · 2023 · github.com