Cherry Studio - Desktop AI Client
🍒 Cherry Studio is a powerful desktop AI client designed for professional users, integrating 30+ industry-specific AI assistants to meet various work scenario requirements and significantly improve work efficiency.
- Official Website: https://www.cherryai.com.cn
- Download: https://www.cherryai.com.cn/download
- Documentation: https://docs.cherryai.com.cn
88API Integration Method
Method 1: One-click import from the 88API console (recommended)
Prerequisite: Cherry Studio is installed and has been opened at least once. Otherwise the browser cannot open Cherry Studio.
- On the API Keys page of the 88API console, click the "⋯" at the far right of the row of the key you want to use, then choose Chat → Cherry Studio.
- When the browser asks whether to open Cherry Studio, allow it.
- Cherry Studio shows an Add 88API API Key dialog. Check that Provider Name is
88APIand Base URL ishttps://api.88api.ai, then click Add. If you see "88API already has this API key. It will not be added again.", the key was imported before; just click Confirm. - Cherry Studio opens the 88API provider under Settings → Model Provider. It is separate from the built-in New API provider and does not change your existing New API settings. If the enable switch in the top-right corner is off, turn it on, then add models as described in step 3 of Method 2 below.
Older versions of the console imported the key into the built-in New API provider (appended after its existing keys). If you imported a key that way before, you can delete the extra key in the New API provider and use the 88API provider from now on.
Method 2: Manual configuration
Parameter Configuration
Provider: the built-in New API provider (Settings → Model Provider → New API)
API Key: Obtain from 88API
API Host: https://api.88api.ai (Cherry Studio appends /v1 or /v1beta automatically depending on the protocol; https://api.88api.ai/v1 also works)
Step-by-Step Guide
-
Copy API key from 88API

-
Add Provider
- Open Settings → Model Provider, select the built-in New API provider in the list, and turn on the enable switch in the top-right corner
- API Key: paste your 88API API Key
- API Host: enter
https://api.88api.ai(Cherry Studio appends/v1or/v1betaautomatically;https://api.88api.ai/v1also works. Do not end it with#, and do not paste/chat/completions)

-
Add Models
Click Sync models at the top right of the model list. In the panel that opens, click + next to a model to add it, or click Add all models to add everything at once. Added models appear in the model list, grouped by series.

-
Return to Chat Page
Click the chat icon in the left sidebar to go back to the chat page.
-
Switch to 88API Model
Click the model name at the top of the chat page and pick a model under the New API group (88API if you used one-click import) to start chatting.

Models with a green eye icon accept images, such as
deepseek-v4.1-flash: select it, click + in the input box to attach an image, and ask your question. No extra setup is needed.
Drawing in Cherry Studio
First, you need to configure an API provider that supports drawing
-
First, click Sync models under the 88API provider (one-click import) or the New API provider (manual configuration), then search for and add the image-generation model
gpt-image-2 -
Click the Paintings icon in the left sidebar, choose
gpt-image-2under 88API or New API in the model selector below the input box, enter a prompt, then click the send button to generate
Using Nano Banana (Gemini image generation)
The Nano Banana models (gemini-3-pro-image, gemini-3.1-flash-image) only work through the native Gemini API. In Cherry Studio 2.0 and later, the chat page and the Paintings page need opposite model settings, so you need two providers:
| Where you generate images | Provider | Model Type |
|---|---|---|
| Chat page | 88API | Text |
| Paintings page | 88API Paint (new) | Image |
When a model's type is Image, the chat page hides it from the model list, while the Paintings page only lists models whose type is Image. The same model cannot serve both pages, so add it once under each provider.
The steps below use the 88API provider created by the one-click import.
Use it in chat
- Open Settings → Model Provider → 88API, click the gear to the right of the API address to open Request configuration, and make sure the Gemini field contains
https://api.88api.ai. Fill it in if it is empty. - Click + in the model list to add a model:
- Model ID:
gemini-3-pro-image(Nano Banana Pro) - Model purpose: Chat
- Chat protocol: Gemini
- Expand More Settings and set Model Type to Text. Turn on Vision under Input Modalities to upload reference images
- Model ID:
- Add
gemini-3.1-flash-image(Nano Banana 2) the same way. - Go back to the chat page, pick Nano Banana under 88API in Select Model, and send your prompt. The generated image appears in the reply.
Use it on the Paintings page
-
Click Add Provider at the bottom of the Settings → Model Provider list:
- Provider Name:
88API Paint(any name works) - Endpoint settings: enter
https://api.88api.aiin the Gemini field - API Key: paste your 88API API key
After adding it, turn on the enable switch in the top-right corner.
- Provider Name:
-
Under this provider, click + to add a model:
- Model ID:
gemini-3-pro-image - Model purpose: Chat (do not choose Image generation)
- Chat protocol: Gemini (if this option is missing, the provider already defaults to Gemini)
- Expand More Settings and set Model Type to Image
- Model ID:
-
Add
gemini-3.1-flash-imagethe same way. -
Quit Cherry Studio completely and open it again. The Paintings page caches its model list; without a restart it may show "Failed to fetch the model".
-
Open the Paintings page, choose Nano Banana under 88API Paint, enter a description, and generate.
Troubleshooting
- "请使用 Gemini 原生接口调用 / please use the Gemini native API": the model purpose is set to Image generation, or the chat protocol is not Gemini. Nano Banana does not support the OpenAI image API; change it to Chat + Gemini.
- Nano Banana is missing on the chat page: its model type under this provider is Image. Change it back to Text.
- Nano Banana is missing on the Paintings page: check that the model type is Image and that you have restarted Cherry Studio. If the model's purpose was ever switched between Image generation and Chat, Cherry Studio removes its image output and the setting cannot be restored in the interface. Delete the model and add it again with the steps above, choosing Chat as the purpose from the start.
- The Paintings page says "Failed to fetch the model": quit Cherry Studio completely and open it again.
CC Switch - All-in-One AI CLI Management Tool
CC Switch tutorial — a cross-platform tool for unified management of provider configs, MCP servers, and system prompts for Claude Code, Codex, Gemini CLI, OpenCode and other AI coding tools. One-click import from 88API.
DeepChat - Commercial-Friendly Full-Featured AI Agent Client
DeepChat guide — connect this open-source, commercial-friendly full-featured AI Agent client to 88API and import DeepChat configuration with DeepLink in one click.