Practical overview
When it fits
Use GPTPortal when you want to self-host a unified chat interface, supply your own provider keys, and work across hosted or compatible local models.
Who it is for
- People who want one interface for models from multiple AI providers.
- Self-hosters who want to chat with models served by compatible local endpoints.
- Developers who want to inspect or modify a lightweight Node.js chat application.
Documented uses
- Switch between configured provider models in one chat interface.
- Send a prompt to multiple models and compare their answers side by side.
- Chat with vision-capable models using images or provide text-based files as prompt context.
- Use voice dictation and spoken replies, or generate images inline.
- Export a conversation as a self-contained HTML file.
Requirements & limits
The application runs on a computer or server controlled by the user. Its documented setup requires a login username and password and at least one provider API key.
- The app requires Node.js 18 or newer and npm; it is not a hosted multi-tenant service.
- Provider credentials are supplied by the user, and model calls go to the selected provider or endpoint.
- Compatible endpoints listed by the project include Ollama, LM Studio and vLLM.
- The README describes a modular Node/Express backend and a hand-written HTML, CSS and vanilla JavaScript frontend, with no build step or database.
- Voice support depends on browser capabilities: the README says Safari works for text but not voice features.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up a local GPTPortal instance
Install and configure the self-hosted portal, then open its chat interface.
Before you start
- Node.js 18 or newer and npm
- A modern browser
- At least one provider API key
- Git, or a downloaded ZIP of the repository
Steps
Clone the repository and enter its folder, or download and extract the ZIP.
Run the command below in the project folder.
npm installStart the server with the command below and open
http://localhost:3000/setup.node server.jsSet a login username and password and enter at least one provider key; save the setup and restart the server.
Open
http://localhost:3000/portal, sign in, choose a model and start a chat.
Expected outcome
A running local portal with the configured provider available for chat.
Guide limitations
- Model requests require an API key for the provider being used.
- Safari supports text but not the documented voice features; the README recommends Chromium browsers or Firefox for voice.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 16 October 2026
- Original website title
- Zaki-1052/GPTPortal · Official homepage title · retained in its source language

