Trae

Add a Trae Custom Model with the correct Full URL behavior, model ID, and vision controls.

Install, configure, and test

1. Install desktop Trae from https://www.trae.ai/download and open a project. Create an API key in the routerra.ai console (https://routerra.ai/keys) and copy the exact model ID. Trae account sign-in does not replace this gateway credential.

2. In the IDE or Agent window, open Settings → Models → Add Model → Custom Model. Choose API Format → OpenAI Chat Completions for this setup. In Custom Request URL, use https://routerra.ai/v1/chat/completions when Full URL is ON. When Full URL is OFF, enter only https://routerra.ai/v1: Trae appends the operation path. Mixing forms can duplicate or omit the path.

3. Set Model ID to YOUR_MODEL_ID and API Key to YOUR_API_KEY. Model Display Name defaults to the ID; changing it creates no gateway alias. In Advanced Configuration, leave defaults unless you know the model's limits. Context Window input/output values must match the selected model, and Supports Image Input should be enabled only for a vision-capable route. Anthropic Messages is a separate API Format, not interchangeable with this Chat Completions URL.

4. Click Add Model. Trae calls the provider interface to validate the key; if it fails, read the provider error logs shown by Trae. After success, enable the model and select it in the chat dropdown. Start a new conversation and check the gateway request log. Restart Trae if the selector remains stale.

If the added model is absent from chat, check its enabled status. For 404 responses recheck Full URL and the exact path; for invalid-model responses compare Model ID with the console rather than the display name.