Docs

1. Point at Cue Cloud

1. Base URL: https://api.cuecloud.io/v1
2. Auth: Authorization: Bearer cue_… from the console.
3. POST /chat/completions with a flagship model id.

Requests carry Authorization: Bearer cue_… from the console. Do not share keys across the team.

2. First request

curl https://api.cuecloud.io/v1/chat/completions \
  -H "Authorization: Bearer cue_••••••••" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v4-pro",
    "messages": [
      { "role": "user", "content": "fix the failing test" }
    ]
  }'

3. Stream

Set stream: true for SSE token chunks. The normal path for IDE agents. Non-streaming completions work for scripts.

curl https://api.cuecloud.io/v1/chat/completions \
  -H "Authorization: Bearer cue_••••••••" \
  -H "Content-Type: application/json" \
  -N \
  -d '{
    "model": "deepseek-v4-pro",
    "stream": true,
    "messages": [
      { "role": "user", "content": "refactor this module" }
    ]
  }'

Models

Open Hub checkpoints we deploy. Specs on models.md.

Clients

Compatible with supported OpenAI-style clients. Tested or documented on the API page: CueCode, Cursor, scripts. Other OpenAI-compatible tools are expected, not claimed here.

• Cursor. Add an OpenAI-compatible provider. Base URL → Cue Cloud. API key → your cue_… key. Model → one of the flagship ids.
• CueCode. Cloud mode on Cue Cloud. Same models and pack include; CueCode can send denser agent signals for this stack.
• Scripts & tools. Any OpenAI SDK or raw HTTP client works. Use it in CI or custom agents — the load we optimize for is still IDE coding.

Errors

Support

[email protected] · API overview · Pricing