Prerequisites
- TRAE IDE installed (trae.ai)
- A ClearMaas API key starting with
sk-
Setup via AI Settings
TRAE supports two protocol paths for custom providers:OpenAI-compatible setup
- Open Settings �?AI (or Trae AI)
- Navigate to Models or Custom Providers
- Add a new provider:
| Field | Value |
|---|---|
| Provider | OpenAI Compatible |
| API Key | sk-... |
| Base URL | https://api.clearmaas.com/v1 |
Anthropic-compatible setup
| Field | Value |
|---|---|
| Provider | Anthropic Compatible |
| API Key | sk-... |
| Base URL | https://api.clearmaas.com |
/v1/messages automatically.
Model naming format
When selecting models in TRAE, use ClearMaas’sprovider/model format:
| Provider | Model ID example |
|---|---|
| Anthropic models | anthropic/claude-sonnet-4.6 |
| OpenAI models | openai/gpt-4o-mini |
| Google models | google/gemini-2.5-pro |
| DeepSeek models | deepseek/deepseek-chat |
| Grok models | grok/grok-4-fast-reasoning |
| Qwen models | qwen/qwen3.6-plus |
| Kimi models | kimi/kimi-k2.6 |
Selecting a model
After adding the provider:- Open the AI chat or Builder panel
- Open the model selector
- Choose ClearMaas (or the provider name you configured)
- Pick any available model
Project-level configuration
For team consistency, you can add model configuration to a project file (if TRAE supports project-scoped AI settings):Environment variable fallback
If TRAE reads standard environment variables, you can set them before launching:Verification
- After adding ClearMaas as a provider, select a ClearMaas model
- Ask a question in the AI chat
- Check your ClearMaas dashboard to confirm the request was routed