Responses
POST/v1/responses
Follows the OpenAI Responses API spec: https://platform.openai.com/docs/api-reference/responses
Request
- application/json
Body
required
- MOD1
- MOD2
Array [
]
- MOD1
Identifier of the model to use.
input
object
required
Input provided to the model.
oneOf
string
string
System-level instructions for the model.
Maximum number of tokens to generate.
Possible values: >= 1
reasoning
object
Reasoning configuration.
Possible values: [low, medium, high]
Whether to stream the response.
falsestream_options
object
Streaming options.
If set, an additional chunk will be streamed with usage statistics
Possible values: <= 2
text
object
Text output configuration.
format
object
Possible values: [text]
Whether multiple tool calls can be executed in parallel.
falsetool_choice
object
Tool selection strategy.
oneOf
string
Possible values: [auto, none, required]
Maximum number of tool invocations.
Available tools for the model.
Possible values: <= 1
Responses
- 200
- 400
Successful Response
- application/json
- text/event-stream
- Schema
- Example (from schema)
Schema
Array [
Array [
]
]
metadata
object
nullable
output
object[]
required
content
object[]
reasoning
object
Possible values: [low, medium, high]
text
object
format
object
usage
object
required
input_tokens_details
object
output_tokens_details
object
{
"id": "string",
"created_at": 0,
"instructions": "string",
"metadata": {},
"model": "string",
"object": "string",
"output": [
{
"id": "string",
"content": [
{
"annotations": [
{}
],
"text": "string",
"type": "string"
}
],
"role": "string",
"status": "string",
"type": "string"
}
],
"parallel_tool_calls": true,
"temperature": 0,
"tool_choice": "string",
"tools": [
{}
],
"top_p": 0,
"max_output_tokens": 0,
"previous_response_id": "string",
"reasoning": {
"effort": "low"
},
"status": "string",
"text": {
"format": {
"type": "string"
}
},
"truncation": "string",
"usage": {
"input_tokens": 0,
"input_tokens_details": {
"audio_tokens": 0,
"cached_tokens": 0,
"text_tokens": 0,
"input_tokens_per_turn": [
0
],
"cached_tokens_per_turn": [
0
]
},
"output_tokens": 0,
"output_tokens_details": {
"reasoning_tokens": 0,
"text_tokens": 0,
"tool_output_tokens": 0,
"output_tokens_per_turn": [
0
],
"tool_output_tokens_per_turn": [
0
]
},
"total_tokens": 0
},
"background": true,
"max_tool_calls": 0,
"service_tier": "string",
"top_logprobs": 0
}
- Schema
- Example (from schema)
- Stream
Schema
- ResponsesStreamResponseCreated
- ResponsesStreamInProgress
- ResponsesStreamOutputItemAdded
- ResponsesStreamReasoningPartAdded
- ResponsesStreamReasoningTextDelta
- ResponsesStreamReasoningTextDone
- ResponsesStreamReasoningPartDone
- ResponsesStreamOutputItemDone
- ResponsesStreamContentPartAdded
- ResponsesStreamOutputTextDelta
- ResponsesStreamOutputTextDone
- ResponsesStreamContentPartDone
- ResponsesStreamCompleted
Array [
Array [
]
]
Array [
Array [
]
]
Array [
]
Array [
]
Array [
Array [
]
]
oneOf
Possible values: [response.created]
response
object
required
metadata
object
nullable
output
object[]
required
content
object[]
reasoning
object
Possible values: [low, medium, high]
text
object
format
object
usage
object
required
input_tokens_details
object
output_tokens_details
object
Possible values: [response.in_progress]
response
object
required
metadata
object
nullable
output
object[]
required
content
object[]
reasoning
object
Possible values: [low, medium, high]
text
object
format
object
usage
object
required
input_tokens_details
object
output_tokens_details
object
Possible values: [response.output_item.added]
item
object
required
A single reasoning or message item in response.output
Possible values: [reasoning, message]
Possible values: [in_progress, completed]
Possible values: [assistant]
content
object[]
Possible values: [reasoning_text, output_text]
Possible values: [response.reasoning_part.added]
part
object
required
Possible values: [reasoning_text]
Possible values: [response.reasoning_text.delta]
Incremental reasoning text fragment
Possible values: [response.reasoning_text.done]
Full accumulated reasoning text for this part
Possible values: [response.reasoning_part.done]
part
object
required
Possible values: [reasoning_text]
Possible values: [response.output_item.done]
item
object
required
A single reasoning or message item in response.output
Possible values: [reasoning, message]
Possible values: [in_progress, completed]
Possible values: [assistant]
content
object[]
Possible values: [reasoning_text, output_text]
Possible values: [response.content_part.added]
part
object
required
Possible values: [output_text]
Possible values: [response.output_text.delta]
Incremental output text fragment
Possible values: [response.output_text.done]
Full accumulated output text for this content part
Possible values: [response.content_part.done]
part
object
required
Possible values: [output_text]
Possible values: [response.completed]
response
object
required
metadata
object
nullable
output
object[]
required
content
object[]
reasoning
object
Possible values: [low, medium, high]
text
object
format
object
usage
object
required
input_tokens_details
object
output_tokens_details
object
{}
data: {"type":"response.created","response":{"id":"resp_AZONkSmI8...","created_at":1788950306,"model":"gpt-oss-20b","object":"response","output":[],"parallel_tool_calls":true,"temperature":1.0,"tool_choice":"none","tools":[],"top_p":1.0,"max_output_tokens":127926,"status":"in_progress","truncation":"disabled","background":false,"service_tier":"auto","presence_penalty":0.0,"frequency_penalty":0.0},"sequence_number":0,"model":"gpt-oss-20b"}
data: {"type":"response.in_progress","response":{"id":"resp_AZONkSmI8...","created_at":1788950306,"model":"gpt-oss-20b","object":"response","output":[],"parallel_tool_calls":true,"temperature":1.0,"tool_choice":"none","tools":[],"top_p":1.0,"max_output_tokens":127926,"status":"in_progress","truncation":"disabled","background":false,"service_tier":"auto","presence_penalty":0.0,"frequency_penalty":0.0},"sequence_number":1,"model":"gpt-oss-20b"}
data: {"type":"response.output_item.added","output_index":0,"item":{"id":"msg_b203b6b1f79704b0","summary":[],"type":"reasoning","status":"in_progress"},"sequence_number":2,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_part.added","content_index":0,"item_id":"msg_b203b6b1f79704b0","output_index":0,"part":{"text":"","type":"reasoning_text"},"sequence_number":3,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":"User","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":4,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":" asks","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":5,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":" simple","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":6,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":" question","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":7,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":": \"What","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":8,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":" is the capital of France?\"","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":9,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.delta","content_index":0,"delta":" Straight answer: Paris.","item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":10,"model":"gpt-oss-20b"}
data: {"type":"response.reasoning_text.done","content_index":0,"item_id":"msg_b203b6b1f79704b0","output_index":0,"sequence_number":22,"text":"User asks simple question: \"What is the capital of France?\" Straight answer: Paris.","model":"gpt-oss-20b"}
data: {"type":"response.reasoning_part.done","content_index":0,"item_id":"msg_b203b6b1f79704b0","output_index":0,"part":{"text":"User asks simple question: \"What is the capital of France?\" Straight answer: Paris.","type":"reasoning_text"},"sequence_number":23,"model":"gpt-oss-20b"}
data: {"type":"response.output_item.done","output_index":0,"sequence_number":24,"item":{"id":"msg_b203b6b1f79704b0","summary":[],"type":"reasoning","content":[{"text":"User asks simple question: \"What is the capital of France?\" Straight answer: Paris.","type":"reasoning_text"}],"status":"completed"},"model":"gpt-oss-20b"}
data: {"type":"response.output_item.added","output_index":1,"item":{"id":"msg_b74061de408cee12","content":[],"role":"assistant","status":"in_progress","type":"message"},"sequence_number":25,"model":"gpt-oss-20b"}
data: {"type":"response.content_part.added","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"part":{"annotations":[],"text":"","type":"output_text","logprobs":[]},"sequence_number":26,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":"The","logprobs":[],"sequence_number":27,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":" capital","logprobs":[],"sequence_number":28,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":" of","logprobs":[],"sequence_number":29,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":" France","logprobs":[],"sequence_number":30,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":" is","logprobs":[],"sequence_number":31,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":" **Paris**","logprobs":[],"sequence_number":32,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.delta","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"delta":".","logprobs":[],"sequence_number":35,"model":"gpt-oss-20b"}
data: {"type":"response.output_text.done","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"text":"The capital of France is **Paris**.","logprobs":[],"sequence_number":36,"model":"gpt-oss-20b"}
data: {"type":"response.content_part.done","item_id":"msg_b74061de408cee12","output_index":1,"content_index":1,"part":{"type":"output_text","text":"The capital of France is **Paris**.","annotations":[]},"sequence_number":37,"model":"gpt-oss-20b"}
data: {"type":"response.output_item.done","output_index":1,"sequence_number":38,"item":{"id":"msg_b74061de408cee12","content":[{"annotations":[],"text":"The capital of France is **Paris**.","type":"output_text","logprobs":null}],"role":"assistant","status":"completed","type":"message"},"model":"gpt-oss-20b"}
data: {"type":"response.completed","response":{"id":"resp_AZONkSmI8...","created_at":1788950306,"model":"gpt-oss-20b","object":"response","output":[{"id":"rs_8728ae8ff9057799","summary":[],"type":"reasoning","content":[{"text":"User asks simple question: \"What is the capital of France?\" Straight answer: Paris.","type":"reasoning_text"}]},{"id":"msg_8be7c8528ed7a59a","content":[{"annotations":[],"text":"The capital of France is **Paris**.","type":"output_text"}],"role":"assistant","status":"completed","type":"message"}],"parallel_tool_calls":true,"temperature":1.0,"tool_choice":"none","tools":[],"top_p":1.0,"max_output_tokens":127926,"status":"completed","truncation":"disabled","usage":{"input_tokens":74,"input_tokens_details":{"cached_tokens":64,"input_tokens_per_turn":[74],"cached_tokens_per_turn":[64]},"output_tokens":37,"output_tokens_details":{"reasoning_tokens":19,"tool_output_tokens":0,"output_tokens_per_turn":[37],"tool_output_tokens_per_turn":[0]},"total_tokens":111},"background":false,"service_tier":"auto","presence_penalty":0.0,"frequency_penalty":0.0},"sequence_number":39,"model":"gpt-oss-20b"}
data: [DONE]
Bad request Error when passing an invalid or non-existing model name or an invalid enum or an invalid type in a field
- application/json
- Schema
- Example (from schema)
Schema
error
object
required
Error returned as response for invalid or missing model name
{
"error": {
"message": "string",
"type": "string",
"param": "string",
"code": "string"
}
}