Setting up another client

Use this page for Cursor, Cline, Aider, Open WebUI, and any other app that can call a custom endpoint. There is no separate guide for those apps. Compare the address the client will call with the four routes below.

The shared order is on Client setup. One Nexus key is enough. The connection entry in the app is not a second Nexus account.

Routes

Protocol	Base URL	Address the preview should show
OpenAI Chat	https://api.nexus-hub.tech/v1	POST /v1/chat/completions
OpenAI Responses	https://api.nexus-hub.tech/v1	POST /v1/responses
Anthropic Messages	https://api.nexus-hub.tech	POST /v1/messages
Gemini-compatible	https://api.nexus-hub.tech/v1beta	/v1beta/models/{model}:generateContent

Read the client's request preview, or the URL it prints before the first call. It must be one of those four. If the preview repeats /v1 or drops it, change the base until the preview is clean. Do not paste the full operation path into the base field.

Fields

• API key: the Nexus key, in a protected field, keychain, or environment variable. Do not add Bearer unless the field asks for the whole header.
• Model: an exact id from GET /v1/models for this key. Gemini names are also listed by GET /v1beta/models.
• Messages version: 2023-06-01 when the client has that field. Nexus requires the header anthropic-version: 2023-06-01 and max_tokens greater than zero.

Check

Send a short message. A normal reply means the connection works.

If it does not work

• 401 or invalid_key: the key is missing, has extra spaces, or was revoked.
• model_not_found or model_not_allowed: the id is not in this key's list.
• unsupported_model: the id is visible, but not on this protocol.
• endpoint_not_allowed: this key cannot use the selected route.
• 404: the preview is not one of the four addresses. Remove the operation path from the base field.

Send support the route, error code, time, and x-request-id. Do not send the key or a full prompt.