Repolode
Submit
AllYourBot

AllYourBot/hostedgpt

A self-hostable Ruby on Rails ChatGPT-style chat app that works with multiple AI providers using your own API keys.

Open-source alternative to:ChatGPT

View on GitHub
Stars
513
Forks
527
Open issues
20
Last commit
9 hours ago

What is it?

HostedGPT is an open-source chat interface modeled on ChatGPT that can connect to OpenAI, Anthropic Claude, Google Gemini, Ollama, Groq and other providers. It is built with Ruby on Rails and can run on a hosted server or a local computer. Users supply their own API keys and pay only for the API usage they generate, rather than paying for separate monthly subscriptions.

The app supports multiple user accounts, so a family or small group can share one instance. A conversation can switch providers or models mid-stream, which makes it practical to compare outputs from different models. Conversations are stored and can be searched across assistants and providers.

Optional features include Google Gmail, Tasks and Calendar tools, web search through a Brave Search API key, file attachments stored in Postgres or Cloudflare R2, and several authentication options including password, Google, Microsoft Graph and HTTP header authentication. Deployment guides cover Render, Fly.io, Kamal, Docker Compose and a bare-metal setup.

Key features

  • 01Connects to OpenAI, Anthropic, Google Gemini, Ollama, Groq and other model providers using your own API keys.
  • 02Switch providers and models within a single conversation to compare answers.
  • 03Supports multiple user accounts for family or group use.
  • 04Search across all saved conversations and providers.
  • 05Installable as a home-screen app on mobile through Safari.
  • 06German localization included.

Use cases

  • Giving a family or small team shared access to several AI models without individual subscriptions.
  • Comparing responses from different LLM providers side by side in one interface.
  • Running a private ChatGPT-style assistant on your own server or laptop with Ollama models.
  • Letting an assistant read and manage Gmail, Google Tasks and Calendar within a chat.

Quick start

  1. 1

    Fork this repository on GitHub using Fork > Create New Fork

  2. 2

    Create a Render.com account and click the Deploy to Render button on your fork, then click Apply

  3. 3

    Open the Web Service URL from the Render dashboard (hostedgpt-XXX.onrender.com)

Things to know

  • The free Render tier reportedly lacks enough memory, so a paid plan (about $7 per month) is probably needed.
  • The free Render service pauses after 15 minutes of inactivity and resumes slowly.
  • Heroku deployment is currently broken according to the README.
  • Fly.io deployment instructions have not been verified in a while.

In collections

Self-hosted appsRun LLMs locally

Topics

#chatgpt#llm#ruby-on-rails#self-hosted#multi-model#ollama#chat-interface

FAQ

What is hostedgpt?
A self-hostable Ruby on Rails ChatGPT-style chat app that works with multiple AI providers using your own API keys. HostedGPT is an open-source chat interface modeled on ChatGPT that can connect to OpenAI, Anthropic Claude, Google Gemini, Ollama, Groq and other providers. It is built with Ruby on Rails and can run on a hosted server or a local computer. Users supply their own API keys and pay only for the API usage they generate, rather than paying for separate monthly subscriptions.
Is hostedgpt free to use? What license does it have?
hostedgpt is open source under the MIT license. Check the license terms before using it in commercial products.
Is hostedgpt still maintained?
Actively maintained. The last commit to hostedgpt was on Nov 7, 2026.
What can I use hostedgpt for?
Giving a family or small team shared access to several AI models without individual subscriptions. Comparing responses from different LLM providers side by side in one interface. Running a private ChatGPT-style assistant on your own server or laptop with Ollama models. Letting an assistant read and manage Gmail, Google Tasks and Calendar within a chat.
How do I install hostedgpt?
Fork this repository on GitHub using Fork > Create New Fork Create a Render.com account and click the Deploy to Render button on your fork, then click Apply Open the Web Service URL from the Render dashboard (hostedgpt-XXX.onrender.com)
Where can I get hostedgpt?
The source code and installation instructions are on GitHub: https://github.com/AllYourBot/hostedgpt

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects