Create a response
Stateless OpenAI-compatible Responses with store forced off and Raon-run web search; vendor storage and other hosted tools are refused.
bearerAuthAuthorizationBearer <token>Raon API key. A Bearer credential takes precedence when both authentication headers are sent.
application/json- body
Stateless Responses request. Nonempty previous_response_id, conversation, or prompt, background: true, item_reference, and uploaded file_id inputs are refused with 400. Send all history inline. Unsupported tool declarations are refused, including declarations inside additional_tools/tool_search_output input items.
model*stringPublic model ID from the model catalog.
1 <= lengthinput*|array<||||||||>instructions?|nullstream?booleanReturn Responses SSE events.
falsemax_output_tokens?integerOutput token cap, including reasoning.
1 <= valuetemperature?|Sampling temperature. Raon models always use temperature=1.0, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
0 <= value <= 2top_p?|Nucleus sampling probability. Raon models always use top_p=0.95, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
0 <= value <= 1top_k?|Top-k sampling cutoff. Raon models always use top_k=20, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
min_p?|Minimum relative token probability. Raon models always use min_p=0, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
presence_penalty?|Topic repetition penalty. Raon models always use presence_penalty=1.5, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
frequency_penalty?|Token frequency penalty. Raon models always use frequency_penalty=0, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
repetition_penalty?|Token repetition penalty. Raon models always use repetition_penalty=1.0, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
tools?array<||||>tool_choice?||parallel_tool_calls?booleantext?reasoning?Raon ignores reasoning effort controls and always enables reasoning. Other reasoning settings are forwarded.
chat_template_kwargs?Serving template options. Raon always enables thinking on all three inference APIs.
store?booleanAccepted, but always overwritten with false, including when omitted. There are no stored responses to retrieve.
falsebackground?booleanOnly foreground execution is supported. true is refused with 400.
include?array<string>Optional backend output inclusions, such as reasoning.encrypted_content. For hosted search Raon removes web_search_call.* inclusions before forwarding and ensures reasoning.encrypted_content is requested.
metadata?Caller metadata passed to the serving backend; this does not enable response storage.
user?stringCaller identifier forwarded to the backend.
truncation?string"auto""disabled"Generated output as JSON, or server-sent events when stream is true. Additional backend fields may be returned.
- response
JSON response from the serving backend; additional backend fields may be present. Raon-run search emits web_search_call items, echoes the caller tools, and reports the usage of its last internal turn, which already holds the whole conversation and every search result.
id*stringobject*stringcreated_at*numberUnix timestamp in seconds.
status?string"completed""failed""in_progress""cancelled""queued""incomplete"model*stringModel ID returned by the serving backend, preserved by the gateway. It can differ from the public ID in the request. Use GET /v1/models IDs for subsequent requests.
1 <= lengthoutput*array<|||||>usage?|nullerror?|nullincomplete_details?|nullinstructions?string|nullmax_output_tokens?integer|nulltools?array<||||>tool_choice?||parallel_tool_calls?booleanstore?booleantemperature?|Sampling temperature. Raon models always use temperature=1.0, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
top_p?|Nucleus sampling probability. Raon models always use top_p=0.95, regardless of supplied values, null or omission. Other providers use their own parameter behavior.
text?reasoning?|nullmetadata?Caller metadata passed to the serving backend; this does not enable response storage.
curl -X POST 'https://raon-api.krafton.ai/v1/responses' \ -H 'Content-Type: application/json' \ -d '{ "model": "raon-1-mini", "input": "Explain why the sky is blue.", "max_output_tokens": 1024, "store": false}'{ "id": "string", "object": "response", "created_at": 0, "status": "completed", "model": "raon", "output": [ { "type": "message", "id": "string", "status": "in_progress", "role": "assistant", "content": [ { "type": "output_text", "text": "string", "annotations": [ { "type": "url_citation", "start_index": 0, "end_index": 0, "url": "string", "title": "string" } ], "logprobs": [ {} ] } ] } ], "usage": { "input_tokens": 0, "output_tokens": 0, "total_tokens": 0, "input_tokens_details": { "cached_tokens": 0 }, "output_tokens_details": { "reasoning_tokens": 0 } }, "error": { "code": "string", "message": "string" }, "incomplete_details": { "reason": "string" }, "instructions": "string", "max_output_tokens": 0, "tools": [ { "type": "function", "name": "string", "description": "string", "parameters": {}, "strict": true } ], "tool_choice": "none", "parallel_tool_calls": true, "store": false, "temperature": 0, "top_p": 0, "text": { "format": { "type": "text" } }, "reasoning": { "effort": "none", "summary": "auto" }, "metadata": { "property1": "string", "property2": "string" }}Create a chat completion POST
OpenAI-compatible chat completion with Raon-run web search, storage disabled, and opt-in streaming usage.
Create a message POST
Anthropic-compatible Messages with Raon-run search and client-executed tools; named tool choice, vendor files, containers and other hosted tools are refused.