Connect an OpenAI-compatible API endpoint
Connect any OpenAI-compatible endpoint
Connect any API that supports the OpenAI chat completions format. This works with any provider that implements the standard chat completions endpoint.
VariesBilling depends on your provider. Some offer free access, others require a paid plan or credits.
How to connect an OpenAI-compatible API
An OpenAI-compatible API accepts the request format used by OpenAI, usually at a base URL ending in /v1. Get the endpoint URL and API key from your provider, then add both under the OpenAI-Compatible provider in MultimodelChat.
- Copy the provider's OpenAI-compatible base URL
- Create or locate the provider API key
- Add the base URL and key in MultimodelChat settings
- Select the exact model IDs returned by the endpoint
Connect it to MultimodelChat
Once you have the required credentials, start your free trial, add this provider in Settings, and save the connection. MultimodelChat encrypts the API key before storing it.
Available models on Multimodel Chat
The models available depend on your endpoint. Any model your provider serves through the OpenAI-compatible chat completions API will work.
Step-by-step guide
Get Your Base URL
Find the endpoint URL from your provider's dashboard or documentation.
- Log in to your provider's dashboard
- Look for the API endpoint or base URL in the documentation
- The URL should point to an OpenAI-compatible endpoint (usually ends with /v1)
- Copy the full URL — you'll need it in the next step
Get Your API Key
Generate or locate your authentication token.
- Navigate to the API Keys or Credentials section in your provider's dashboard
- Create a new key or copy an existing one
- Copy the key immediately — some providers only show it once
Store your key securely. If you lose it, you may need to revoke and regenerate it.
Add to Multimodel Chat
Connect your endpoint in Multimodel Chat settings.
- In Multimodel Chat, go to Settings → API Keys
- Select OpenAI-Compatible
- Paste your Base URL
- Paste your API Key
- Click Save
Select Your Models
Choose which models from your endpoint to use.
- After saving, Multimodel Chat loads the model list from {your-base-url}/models
- If the URL is missing the /v1 suffix, it is detected and fixed automatically
- If your server doesn't expose a /models list, enter the exact model ID manually in the provider form
- You can switch between models mid-conversation
Common issues
Swipe horizontally to see the full table.
| Issue | Fix |
|---|---|
| Connection refused | Check the Base URL is correct and the server is running |
| 404 Not found | The path before /models matters — usually ending in /v1. If the first request misses it, the app retries with /v1 appended automatically |
| No models returned | Some servers don't expose a /models list — add the exact model ID manually in the provider form |
| Authentication failed | Some providers use Bearer token format — check their docs |
| Model not found | Ensure the model name matches exactly what your provider serves |
| CORS errors (local) | Ensure CORS is enabled on your local server |
Frequently asked questions
What does OpenAI-compatible mean?
It means the API accepts the same request format as OpenAI's API (POST to /v1/chat/completions with the same JSON structure). Many providers support this format, making them easy to integrate.
My base URL doesn't end in /v1 — is that wrong?
Not necessarily. Multimodel Chat calls {baseURL}/models, so the URL must include every path segment before /models. If you paste the root domain and models fail to load, the app automatically retries with /v1 appended.
Can I use a locally hosted model?
Yes. If your local server exposes an OpenAI-compatible endpoint and Multimodel Chat can reach it, you can connect it as a custom provider. You may need to expose the endpoint via a tunnel if Multimodel Chat is cloud-hosted.
Set up another provider
Related guides
Provider requirements and pricing last reviewed August 5, 2026. Check the linked provider documentation for current terms before purchasing credits.
Ready to connect your OpenAI-Compatible key?
Paste your API key and Multimodel Chat tests it live against the provider before saving. Choose hosted encrypted storage or keep the key in your current browser profile.
Start the 7-day free trial