API

Responses API

POST /v1/responses accepts the OpenAI Responses API format. Use it if your code or SDK already targets Responses.

Request

curl https://api.llm-hub.dev/v1/responses \
  -H "Authorization: Bearer $LLMHUB_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.6-sol",
    "instructions": "Answer briefly.",
    "input": "Name three rivers in Europe.",
    "max_output_tokens": 200
  }'
Python (openai)
response = client.responses.create(
    model="gpt-5.6-sol",
    input="Name three rivers in Europe.",
)
print(response.output_text)

Response

The response is a response object with status, output items, and usage (input_tokens, output_tokens, total_tokens). When max_output_tokens is reached, status is incomplete and incomplete_details.reason is max_output_tokens.

With "stream": true you receive typed events such as response.created, response.output_text.delta, and a final response.completed or response.incomplete.

Last updated