Use DeepSeek and GLM in Cherry Studio
Updated 2026-10-18 · Add an OpenAI provider to the Cherry Studio desktop app: API host, key and model ID, with the fix for a 404.
Connection details
- Base URL (OpenAI compatible)
- https://api.aifans.fan/v1
- API key
- Create one in the console; it starts with sk- (no account yet? sign up)
- Recommended models
- deepseek-v4.1-flash, deepseek-v4-pro (all models and prices)
Cherry Studio is a desktop AI client for Windows, macOS and Linux: chats, assistants, a knowledge base and translation in one app. It connects to any OpenAI-compatible service, so every model here works in it.
Add the provider
- Settings (the gear icon) → Model Provider → Add, name it AIFans API and set the provider type to OpenAI.
- API key: the key from the console.
- API host: https://api.aifans.fan — Cherry Studio appends /v1/chat/completions itself. If a check comes back 404, use https://api.aifans.fan/v1/ instead, with the trailing slash that tells it to take the address as given.
- Add model: the model ID deepseek-v4.1-flash (and deepseek-v4-pro for harder work); the models page lists them all.
- Switch the provider on, then run its check or send a message in a new chat.
The provider that ships with the app has a fixed address: add your own instead. Menu names differ a little between versions.
Good to know
- Cherry Studio's knowledge base needs an embedding model, which we do not offer; take that from another provider and keep the chat models here.
- Where a model thinks before answering, the thinking is billed as output tokens. For small talk and translation a model without thinking is cheaper — the models page says which is which.
- When several people share one install, give each of them a key with its own daily cap; the console shows usage per key.