Tool Integration

info

The page and service rules are subject to the current Console.

Most of the clients / tools below support the OpenAI-compatible protocol — just configure the Base URL and API key and you are ready, no code required. Field names may vary slightly between tools and versions; the examples below use the common versions.

First, prepare these two items

Setting	Value
Base URL	https://llmapi.cttai.art/v1
API Key	Create one on the API Console "Tokens" page

Two general rules: do not append /v1 again (e.g. .../v1/v1 is wrong); paste the API key in full without extra spaces or newlines.

Cherry Studio (desktop client)

1. Open Settings → Model Service → Providers.
2. Click "Add provider" and choose type OpenAI (OpenAI-compatible).
3. Fill in:
   • API Domain / API address: https://llmapi.cttai.art/v1
   • API Key: your API key
4. Click "Check" or add a model manually, select it and start chatting.

Models not in the default list can be added manually by entering their model ID (see the Model Marketplace).

ChatBox (desktop / mobile)

1. Open Settings → Models.
2. Add a custom provider (AI provider) with type OpenAI.
3. Fill in:
   • API Host: https://llmapi.cttai.art (ChatBox usually appends /v1 automatically; if your version does not, use https://llmapi.cttai.art/v1)
   • API Key: your API key
4. Add a model (you can add a model ID manually), save it and select it to chat.

NextChat (ChatGPT-Next-Web) (web / self-hosted)

Option A: self-hosted environment variables

OPENAI_API_KEY=your-api-key
OPENAI_BASE_URL=https://llmapi.cttai.art/v1

Option B: in-app settings

1. Open Settings → Model Provider → Custom endpoint.
2. Endpoint: https://llmapi.cttai.art/v1
3. API Key: your API key
4. Add the model ID under "Custom models" and start chatting.

LobeChat (web / desktop)

1. Open Settings → Model Providers → OpenAI.
2. Fill in:
   • API Key: your API key
   • Proxy URL (OpenAI Proxy URL): https://llmapi.cttai.art/v1
3. Enable the target models and start chatting.

For the self-hosted version you can also use the env var OPENAI_PROXY_URL=https://llmapi.cttai.art/v1.

one-api / other relay panels (server-side panel)

Add this service as an upstream channel in your own relay panel:

1. In the panel go to Channels → Add channel.
2. Choose type OpenAI.
3. Channel address / Base URL: https://llmapi.cttai.art/v1
4. Key: your API key
5. Configure groups, allowed models and quotas as needed, save and "Test" the channel.

Verification and troubleshooting

• Send a small request in a new conversation (e.g. "Please just reply OK"), then check the Console "Usage Logs" for time and model.
• 401: check auth — is the key correct, free of extra spaces / newlines, and not deleted?
• 404: check the address and protocol — did you append /v1 twice, or misspell the model ID?
• 403: check whether the key's group permissions include the target model.
• Model list empty or missing a model: query the models visible to the key via /v1/models, or enter the model ID manually.