Setup guide

Provider setup guide

Browse setup steps and troubleshooting by tool.

Provider setup guide/LobeChat Configuration

LobeChat Configuration

LobeChat is the web chat tool of LobeHub. This is about integrating OpenAI or OpenAI-compatible providers into LobeChat: first get the API Key, Base URL, and model name from the provider's backend, and then fill them into LobeChat's Settings - Language Model.

Official information Downloads and documentation are subject to the official site.
Official Website Understanding the LobeHub / LobeChat ecosystem Official documentation Model Service and Self-Deployment Instructions GitHub Open source repository and deployment entry

First understand what this page is doing

The official documentation places the configuration entry in Settings inside Language Model. You can think of it as a 'model provider management page': after enabling the OpenAI provider, LobeChat will know where to send requests, which Key to use, and which models appear in the chat box.

If your provider is an AI provider, you can still choose OpenAI on the LobeChat page. The only difference is API Key and API Proxy Address. You need to fill in your own AI provider information, not the OpenAI official default address.

Set Entry Reference First go to the language model settings page, then expand the OpenAI provider. The screenshot below is only for confirming the entry and field area.
Language Model settings in the official LobeChat documentation

How should the several fields be filled in?

The screenshot is from the official LobeChat documentation. The interface in the screenshot is in English, so the original English field names are retained below; the names you see in the Chinese interface may be 'Language Model', 'API Key', 'API Proxy Address', 'Model List'.
Configuration itemFilling Method
OpenAI switchFirst, open this provider. If the switch is not turned on, even if the Key is filled in, the corresponding model may not be visible on the chat page.
API KeyEnter the Key you created in the provider's backend. It may also be called a token, Key, or secret key. Do not use the example content from the screenshot.
API Proxy AddressThis is the Base URL / API link. The official website's default example is https://api.openai.com/v1; the AI provider needs to be changed to the API link given to you by the backend, some include /v1, some do not.
Use Client-Side Fetching ModeOrdinary users should first follow the provider's tutorial instructions. If the provider indicates that a client request is required, then enable it; if you have a self-built deployment and the server forwards it uniformly, it usually does not need to be enabled.
Model ListWhich models can be selected on the chat page is determined here. Give priority to obtaining the model list; if retrieval fails, manually add them according to the model names in the provider's backend.
Connectivity CheckUsed to check whether the Key and API Proxy Address can connect. If the check fails, first check the Key, Base URL, and model permissions.

Configuration Steps

  1. Open LobeChat and go to Settings - Language Model. If the interface is in English, go to Settings - Language Model.
  2. Expand the OpenAI section and turn on the switch on the right.
  3. In API Key, enter the Key generated by the provider backend.
  4. In API Proxy Address, enter the Base URL / API link provided by the provider backend. Do not guess whether you need to add it yourself /v1, follow the backend example.
  5. Click Get Model List to get the model list; if the retrieval fails, manually add the backend model name to Model List.
  6. Click Check to run a connectivity check. After passing, go back to the chat page and select the newly configured model from the model dropdown.
Model selection reference After configuration, return to the chat page to confirm that the model you just configured can be selected in the model dropdown.
Model selection on the LobeChat chat page

Self-hosted deployment environment variable example

LobeChat OpenAI compatible environment variable example
OPENAI_API_KEY=YOUR_API_KEY OPENAI_PROXY_URL=YOUR_PROVIDER_API_URL
When self-deploying, the names of the environment variables may change with the version. The page keeps the official documentation entry; the real fields are subject to your current version of the documentation.