Build

Endpoints

The three request formats Kael speaks, the routes for each, and what comes back.

Routes

Kael speaks three request formats. Pick the one your SDK or framework already uses; the answer and the price are the same in all three.

Route	Format
POST`/v1/chat/completions`	OpenAI Chat Completions
POST`/v1/responses`	OpenAI Responses
POST`/v1/messages`	Anthropic Messages
GET`/v1/models`	Lists the models you can call

OpenAI SDKs use the base URL https://api.quancis.space/v1 and add /chat/completions or /responses themselves.
Anthropic SDKs use https://api.quancis.space with no /v1, because they add /v1/messages themselves.
One API key works on every route. See Authentication.

Model name

Send kael-beta as model. Any other value returns a 400 with the code model_not_found. The older names quan-pro and quan-4 still resolve, but they are deprecated.

GET /v1/models

{
  "object": "list",
  "data": [
    {"id": "kael-beta", "object": "model", "created": 1755000000, "owned_by": "quancis"}
  ]
}

The same request in each format

POST /v1/chat/completions

{
  "model": "kael-beta",
  "messages": [
    {"role": "system", "content": "You are a careful code reviewer."},
    {"role": "user", "content": "Review this function for bugs: ..."}
  ],
  "reasoning_effort": "high"
}

Item	Chat Completions	Responses	Messages
System prompt	a `system` or `developer` message	`instructions`	top-level `system`
Thinking level	`reasoning_effort`	`reasoning.effort`	`reasoning_effort`
Answer text	`choices[0].message.content`	the `message` item in `output`	the `text` block in `content`
Tokens billed	`usage.prompt_tokens`, `usage.completion_tokens`	`usage.input_tokens`, `usage.output_tokens`	`usage.input_tokens`, `usage.output_tokens`
Cached tokens	`usage.prompt_cache_hit_tokens`	`usage.input_tokens_details.cached_tokens`	Not reported
Why it stopped	`finish_reason`: `stop`, `length`, `tool_calls`	`status`: `completed`, `incomplete`	`stop_reason`: `end_turn`, `max_tokens`, `tool_use`

Response shapes

The Chat Completions response is shown in the Quickstart. The other two look like this. A Responses output list can hold more than one kind of item, so read the one whose type is message.

{
  "id": "...",
  "status": "completed",
  "model": "kael-beta",
  "output": [
    {
      "type": "message",
      "role": "assistant",
      "content": [
        {"type": "output_text", "text": "The range goes one step too far: ..."}
      ]
    }
  ],
  "usage": {
    "input_tokens": 38,
    "output_tokens": 214,
    "total_tokens": 252,
    "input_tokens_details": {"cached_tokens": 0}
  }
}

The model field always says kael-beta. The usage you see is what you are billed. See Pricing and billing.