Base URL and Model Selection
CertaEdge's website address and model API paths serve different purposes. Before entering an address, check whether your client expects the service root URL, an SDK Base URL, or a full endpoint.
Use the address your protocol expects
| Use case | Address |
|---|---|
| OpenAI-compatible SDKs, Dify, and n8n credentials | https://certaedge.com/v1 |
| Cherry Studio 2.1.10 custom provider's OpenAI endpoint | https://certaedge.com |
Claude Code's ANTHROPIC_BASE_URL | https://certaedge.com |
| A direct Chat Completions HTTP request | https://certaedge.com/v1/chat/completions |
| A direct Responses HTTP request | https://certaedge.com/v1/responses |
Clients may append /v1 or an endpoint name automatically. If you receive HTML, or the path contains /v1/v1, check for missing or duplicate path segments.
Choose a model
- Open the model catalog.
- Confirm that the model supports the protocol and features you need, such as tool calling, vision, or streaming.
- Copy the full model ID, preserving its capitalization, hyphens, and version suffix.
- Confirm that your API Key's group and model restrictions allow access to it.
A model's display name can differ from its API model ID. Copy the ID provided on the page rather than using a custom chat name. First choose a model that supports your target protocol, then enable that same protocol in your client.
If the catalog does not show supported protocols, contact platform support before proceeding. A model ID returned by the model list endpoint only confirms that it can be listed, not that it supports every endpoint.
Next step
Complete one test request, then open the tool guides to configure your client.