Windsurf IDE with Meridian
Use Meridian as a custom provider in Windsurf IDE to access GPT-5 Codex, GPT-4.1, and DeepSeek V4 Pro directly inside Cascade.
Add Meridian as a Custom Provider
- Open Windsurf and navigate to Settings → Cascade → Custom Provider.
- Set Provider to
OpenAI Compatible. - Set Base URL to
https://meridian.getnimbus.net/api/v1. - Paste your Meridian API key (format:
nim_live_***) into the API Key field. - Under Models, add one or more of the recommended models listed below.
Recommended Models
| Model ID | Best for |
|---|---|
gpt-5-codex | Complex refactors, architecture, multi-file edits |
azure-swc/gpt-4.1 | Fast completions, inline suggestions |
azure-swc/DeepSeek-V4-Pro | Long-context analysis, large codebase reasoning |
Verifying the Connection
After saving, open Cascade and send a test prompt. If the model responds, your Meridian integration is active. Check the Usage dashboard to confirm tokens are being counted against your Nimbus plan.
Meridian is a Nimbus product. All requests are routed through Nimbus infrastructure and subject to your plan's rate limits and data retention policies. API keys are scoped per workspace and can be rotated from the API Keys page.