Run the EveryAPI gateway inside LobeChat
Configure EveryAPI as an OpenAI-compatible endpoint in LobeChat, then choose from the models available to your account. Availability, latency, and protocol-specific behavior depend on the selected model and current gateway routes.
One base URL line points LobeChat at EveryAPI
Default global Base URL / China-first Base URL: https://api.everyapi.ai/v1 / https://api-cn.everyapi.ai/v1
# 1. LobeChat → Settings → Language Model# 2. Open the OpenAI-compatible provider settings API Key sk-everyapi-...API Proxy URL https://api.everyapi.ai/v1Model List add IDs available in your EveryAPI account # 3. Save and test each model and feature you plan to useFrom zero to your first call
- Install LobeChat (skip if already installed)
Three options: the online version at chat.lobehub.com, the Mac/Win desktop app, or self-host via Docker (one-line docker compose).
lobehub.com/chat
- Open Settings → Language Model
Click the gear icon in LobeChat's top-right to open Settings, then select "Language Model" in the left menu.
- Change two fields on the OpenAI entry
Set API Key to your EveryAPI sk-everyapi-... key and API Proxy URL to https://api.everyapi.ai/v1. Click "Check connectivity" and look for the green checkmark.
- Select / fetch the model list
In the "Model List" section on the same page, click "Get model list" and LobeChat automatically pulls the live model catalog from EveryAPI. You can also manually check the few you want to appear in the picker.
- Start chatting
Return to the main view and choose a model fetched from your EveryAPI account. Test the model and features you need before relying on the configuration.
LobeChat × EveryAPI
Can Claude / Gemini models be used in this OpenAI entry too?
Yes. At the protocol level, the EveryAPI gateway automatically translates OpenAI Chat Completions requests into Claude Messages / Gemini generateContent — LobeChat sends requests with the OpenAI client and you just set the model name to claude-sonnet-5 / gemini-3.1-pro-low.