Chat Completions API

Sign in and open Dashboard / API to generate or revoke your key. Keys are shown only once after generation.

If a key is exposed in logs, screenshots, chat, or source code, regenerate it immediately. Revoked keys stop working for all integrations.

Use either OpenAI-style bearer auth or the direct API key header.

Authorization: Bearer wrees_your_api_key
x-api-key: wrees_your_api_key

The API accepts the request's model field only when it matches an allowed Wrees model and that model is enabled for your API key by an admin.

When the upstream is configured for OpenRouter, model IDs ending in :free are also accepted and use the same local request and token limits.

Admins set each user's input tokens per request, output tokens per request, daily input cap, and daily output cap. Input tokens are counted before the upstream call. Output tokens are capped before forwarding the request when possible.

Streaming is not currently supported. Requests with stream: true return a 400 response.

curl https://wrees.com/v1/chat/completions   -H "Authorization: Bearer wrees_your_api_key"   -H "Content-Type: application/json"   -d '{
    "model": "gpt-5.4-mini",
    "messages": [
      { "role": "user", "content": "Write a one-line product tagline." }
    ],
    "max_completion_tokens": 80
  }'

const response = await fetch("https://wrees.com/v1/chat/completions", {
  method: "POST",
  headers: {
    "Authorization": `Bearer ${process.env.WREES_API_KEY}`,
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
    model: "gpt-5.4-mini",
    messages: [
      { role: "system", content: "You are concise." },
      { role: "user", content: "Explain rafters in one sentence." }
    ],
    max_completion_tokens: 120
  })
})

const data = await response.json()
console.log(data.choices[0].message.content)

Successful responses are relayed from the upstream AI provider in the OpenAI chat completion format.

{
  "id": "resp_...",
  "object": "chat.completion",
  "model": "gpt-5.4-mini",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 42,
    "completion_tokens": 24,
    "total_tokens": 66
  }
}

If your request fails, check that your key is active, the model is supported, and your token usage is under the configured limits. For account help, contact support.