Windsurf is Codeium’s IDE; Cascade is its inline assistant. Windsurf’s public docs clearly document BYOK and proxy settings, but they were not specific enough here to verify a stable field-by-field custom base URL flow. Use the steps below as an in-product setup guide, and confirm labels against your current Windsurf build. Official references used:Documentation Index
Fetch the complete documentation index at: https://docs.routing.run/llms.txt
Use this file to discover all available pages before exploring further.
Setup
Get your API key
Create an API key at app.routing.run. Copy the key — it starts with
rk_.Open Windsurf settings
Open Windsurf and press
Cmd+, (macOS) or Ctrl+, (Windows/Linux). Navigate to Cascade settings.Add routing.run as a custom provider
Select the OpenAI-style or custom provider flow if your Windsurf build exposes one. Name the provider
routing.run if the UI offers a provider name field. Then paste your dashboard secret (rk_…) as the API key and set the base URL to https://api.routing.run/v1 with no trailing slash.If the UI hides advanced fields, look under Cascade, BYOK, or custom provider settings. Windsurf’s public docs were clearer about BYOK than about raw base URL field names.
Manual configuration
If you want to fill Windsurf settings manually, use these exact values:| Field | Value |
|---|---|
| Provider name | routing.run if the UI offers a name field |
| Integration type | OpenAI |
| Base URL | https://api.routing.run/v1 |
| API Key | rk_REPLACE_ME |
| Model ID | route/glm-5.1 |
route/ prefix on model ids. If Windsurf lets you save multiple providers, name this one routing.run so you can switch between providers more easily.
I could not verify a stable public field-by-field custom endpoint reference for Windsurf from official docs alone. If your build uses different labels, match the same underlying values: API key, base URL, and
route/… model id.Recommended models
Published model IDs are listed on the models page. Exact access depends on your plan tier and the dashboard; common picks:| Model | Use case |
|---|---|
route/glm-5.1-precision | Strongest reasoning-first choice |
route/qwen3.6-plus | Best default starting point |
route/minimax-m2.7-highspeed | Long-context sessions |
route/kimi-k2.5 | Agentic and tool-heavy flows |
route/glm-5.1 | Faster GLM alternative |
route/qwen3.5-plus | Qwen3.5 flagship |