Skip to main content
Cherry Studio is an open-source cross-platform desktop AI client. It can hold configs for several LLM providers at once, supports a local knowledge base and agents, and connects to RuAPI through the OpenAI-compatible protocol.

Prerequisites

1

RuAPI account + API key

Console → API Keys → create a sk-... key.
The quickest way is to take ready-made settings from the Console. Open Console → API Keys, click Connect Claude Code, Cline & more and pick Cherry Studio. The Add RuAPI to Cherry Studio button adds the RuAPI provider for you (if Cherry Studio is starting for the first time, click Set up later at the top right of the welcome screen first), and the window also has fields with your key for manual setup. Install the app as shown below and follow the steps in the window (provider, the deepseek-flash model, check); then, in “Add RuAPI as a provider”, only the last step is left: choosing default models.
RuAPI API Keys page: the https://www.ruapi.ai address and the “Connect Claude Code, Cline & more” button

Console → API Keys: the “Connect Claude Code, Cline & more” button

The “Connect an app” window in the RuAPI Console with ready-made settings for Cherry Studio

The “Connect an app” window with Cherry Studio selected: the address and key are filled in (key hidden in the screenshot)

Install Cherry Studio

Download from cherry-ai.com or GitHub Releases:
Download Cherry-Studio-Setup-x.y.z.exe and run the installer.

Add RuAPI as a provider

These steps and screenshots are for Cherry Studio 2.1.6 (October 2026). Button names may differ slightly in other versions.
1

Open the add-provider form

On first launch, click Set up another provider, then Add Provider at the bottom of the list.If Cherry Studio is already set up: Settings → Model Provider → Add Provider.
2

Fill in the form

Below the address, Cherry Studio shows the request path https://www.ruapi.ai/v1/chat/completions, which means the address is right. You can leave the Anthropic field empty. Click Add.
Cherry Studio: add-provider form for RuAPI with the address https://www.ruapi.ai

The new provider form: name, key and OpenAI address

3

Add the deepseek-flash model

Cherry Studio opens the RuAPI model list — click Skip in it. Then click + to the right of Sync models, enter deepseek-flash in Model ID (Model Name and Group Name fill in with the same value automatically) and click Add Model.
If you haven’t topped up yet, the trial $0.05 (added to new accounts at sign-up) only works with deepseek-flash, claude-haiku-5-5 and gpt-6-luna. Once you top up $5 or more, add other models with Sync models — just tick them in the list. Prices for every model are on the pricing page.
Cherry Studio: Add Model window, model ID, name and group are deepseek-flash

Adding the model: Model ID is deepseek-flash

4

Check and enable

Click Model Check to the right of the API key, then Start in the window that opens. When the check passes, the switch at the top right turns green — the provider is enabled.
Cherry Studio: check window with deepseek-flash selected and the Start button

Checking deepseek-flash

The provider page shows the key, the address and the deepseek-flash model.
Cherry Studio: RuAPI provider page with key, address and the deepseek-flash model

The RuAPI provider is set up and enabled

5

Choose default models

On first launch, click Next at the bottom right to open Choose Default Model. In the first drop-down, pick deepseek-flash under RuAPI — the other two fields fill in with the same model. Click Get Started.
Cherry Studio: default assistant, quick and translate models are all deepseek-flash from RuAPI

All three default models are deepseek-flash

If Enable the API Gateway? pops up next, click Cancel: the gateway is only needed for Cherry Studio agents; regular chat works without it.

First call

Open Chat (the speech-bubble icon on the left), click the model name at the top and pick a model under RuAPI. Send:
A normal reply plus a matching RuAPI Console → Usage Logs entry = success.

Default model choices

Settings → Default Model:
  • Default Assistant Model: everyday chat. Recommend the GPT family or Claude Sonnet.
  • Quick Model: simple tasks such as naming conversations and extracting search keywords. A cheap model like Claude Haiku is enough.
  • Translate Model: GPT mini or Claude Haiku work well.
While you’re on the trial balance, pick deepseek-flash under RuAPI in these fields: by default Cherry Studio uses its own built-in model, not RuAPI.

Troubleshooting

  • The API key must be a RuAPI key (sk-...), not an OpenAI key.
  • Watch for trailing whitespace.
  • Check the key in RuAPI Console → API Keys.
  • Model ID typo. When you add a model by hand, Model ID must match RuAPI exactly (for example deepseek-flash), with no spaces before or after. For other models, it’s easier to click Sync models and tick them in the list.
  • Don’t append /chat/completions to the API host — https://www.ruapi.ai is enough.
Streaming (SSE) is enabled by default on both sides. If it stutters:
  • Check network stability.
  • Try disabling streaming in settings to compare.
Cherry Studio’s MCP relies on model-level function calling. GPT and Claude models handle it reliably — see function calling.
Make sure the selected model supports vision. On RuAPI, GPT, Claude and Gemini models and deepseek-flash understand images; GLM models are text-only — see vision.
Cherry Studio supports any number of providers. Add each one the same way. The model dropdown in the top-right of each chat switches between them.

Advanced

  • Knowledge base: Cherry Studio’s built-in RAG needs an embedding model. RuAPI offers text-embedding-3-small and text-embedding-3-large: add one to the RuAPI provider with the ”+” button like any other model (by ID; if Cherry Studio does not detect the type itself, mark it as an embedding model), then pick it when you create the knowledge base. Embedding models unlock after a top-up. If your documents must not leave your computer, use a local embedding model (for example via Ollama) and keep chat on RuAPI.
  • Agents: Cherry Studio’s agent system runs client-side and calls your configured provider. The Claude Sonnet or GPT families make solid agent backbones.
  • Multiple keys: if one key isn’t enough, create several with separate limits in the RuAPI Console and bind them to different Cherry Studio profiles.