Chat with a local AI model running in Ollama or LM Studio from a simple browser page.
Compare answers from two different AI models side by side.
Try out a cloud AI provider's API key without installing any software.
| immrshervin/one-chat | 1000sbin/moa-goods-maker | brunosimon/esin-e1-p2026 | |
|---|---|---|---|
| Stars | 11 | 11 | 11 |
| Language | HTML | HTML | HTML |
| Last pushed | — | — | 2022-05-15 |
| Maintenance | — | — | Dormant |
| Setup difficulty | easy | easy | easy |
| Complexity | 1/5 | 1/5 | 1/5 |
| Audience | vibe coder | general | general |
Figures from each repo's GitHub metadata at analysis time.
Just download the HTML file and open it in a browser, cloud providers need their own API key.
One Chat is a chat interface for talking to AI models, both ones running on your own computer and ones running in the cloud. Instead of switching between a different app or website for each AI provider, you open one page and connect it to whichever model you want to use. The whole application is a single HTML file, so there is nothing to install, no build step, and no server needed to run it. You just download the file and open it in a browser. It works out of the box with many providers, including local tools like Ollama and LM Studio, and cloud services like OpenAI, Google Gemini, DeepSeek, Groq, and several others, plus any service that offers an OpenAI compatible API. Local providers usually do not need an API key, while cloud providers do. Once connected, you get streaming responses that appear as they are generated, formatted text with tables and code blocks, buttons to copy responses or code, a mode to compare answers from two models side by side, and settings to adjust the system prompt, temperature, and maximum response length. You can also upload text files, JSON, Markdown, and images, save and search your chat history, export conversations, back up and restore your data, and switch between English and Persian interfaces with light and dark themes. Getting started means downloading the HTML file, opening it in a browser like Chrome, Edge, or Firefox, choosing your provider in the settings, entering the endpoint and model name and API key if one is required, testing the connection, and then chatting. Connecting to a local tool like Ollama or LM Studio just means pointing it at the local address those tools run on and picking the model you already have loaded there. On privacy, everything is stored locally in your browser and requests go directly from your browser to whichever provider you chose, without passing through any server run by the project. Your API key can be kept only for the current browser session or saved in local storage depending on a setting, and the README recommends not saving it on shared computers and leaving it out of backups for better security. The project describes itself as a simple, lightweight alternative rather than a replacement for larger, full featured AI client applications.
A single HTML file chat app that connects to local or cloud AI models with no install, build step, or server required.
Mainly HTML. The stack also includes HTML, JavaScript.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
double-check against the repo, no cap.