API

Chat Completions

POST /v1/chat/completions is the main text generation endpoint. It accepts the standard OpenAI Chat Completions request and returns the standard response.

Request

curl https://api.llm-hub.dev/v1/chat/completions \
  -H "Authorization: Bearer $LLMHUB_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.6-sol",
    "messages": [
      {"role": "system", "content": "Answer in one sentence."},
      {"role": "user", "content": "What is idempotency?"}
    ],
    "max_tokens": 200
  }'
FieldDescription
modelModel ID from Models. Required.
messagesConversation with system, user, assistant, and tool roles. Required.
max_tokens / max_completion_tokensMaximum output tokens.
stopString or array of strings that end generation.
streamReturn tokens as Server-Sent Events. See Streaming.
toolsFunctions the model may call. See Tool calling.
response_formatJSON output. See Structured output.

Response

{
  "id": "chatcmpl-...",
  "object": "chat.completion",
  "created": 1790675901,
  "model": "gpt-5.6-sol",
  "choices": [{
    "index": 0,
    "message": {"role": "assistant", "content": "...", "refusal": null, "annotations": []},
    "logprobs": null,
    "finish_reason": "stop"
  }],
  "usage": {
    "prompt_tokens": 24,
    "completion_tokens": 19,
    "total_tokens": 43,
    "prompt_tokens_details": {"cached_tokens": 0, "audio_tokens": 0},
    "completion_tokens_details": {"reasoning_tokens": 0, "audio_tokens": 0,
      "accepted_prediction_tokens": 0, "rejected_prediction_tokens": 0}
  },
  "service_tier": "default",
  "system_fingerprint": null
}

finish_reason is stop, length (hit max_tokens), or tool_calls. You are billed for the tokens reported in usage.

Last updated