Skip to main content
POST
/api/v1/responses

Autorisations

Authorization
string
header
requis

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Corps

application/json

Request body for the Responses API endpoint. E2EE-capable models are not supported on /api/v1/responses; use /api/v1/chat/completions with the required E2EE headers instead.

model
string
requis

The ID of the model to use. E2EE-capable models are not supported on /api/v1/responses; use /api/v1/chat/completions with the required E2EE headers instead.

Exemple:

"zai-org-glm-5-1"

input
requis

The input to generate a response for. Can be a simple string or an array of messages.

instructions
string | null

System instructions for the model, applied before input.

include
string[]

Additional response fields to include (OpenAI-compatible).

max_output_tokens
integer

Maximum number of tokens to generate.

Plage requise: x > 0
temperature
number

Sampling temperature between 0 and 2.

Plage requise: 0 <= x <= 2
top_p
number

Nucleus sampling parameter.

Plage requise: 0 <= x <= 1
fallbacks
object[]

Anthropic beta parameter for Claude Fable 5 server-side refusal fallback. Forwarded only for direct Anthropic routes; ignored for other providers.

Maximum array length: 10
Exemple:
reasoning
Reasoning Configuration · object | null

Reasoning controls for models that support reasoning.

parallel_tool_calls
boolean

Whether the model may call more than one tool in a single turn.

tools
(Function Tool · object | Tool Definition · object | Web Search Tool · object | X Search Tool · object | Code Interpreter Tool · object | File Search Tool · object | Computer Use Tool · object | Generic Tool · object)[]

A list of tools the model may call.

tool_choice

Controls which tool is called by the model.

Options disponibles:
auto

Enable web search for this request.

stream
boolean

Whether to stream back partial progress.

anon_user_id
string

Optional identifier for the API customer's end user. Combined with the Venice user id when attributing the request to upstream providers. Distinct from the discarded OpenAI user field. Must be printable ASCII and must not contain ||.

Required string length: 1 - 128
Pattern: ^[\x20-\x7E]+$
Exemple:

"end-user-123"

venice_parameters
Venice Parameters · object

Venice-specific options, such as web search, characters, and the Venice system prompt.

Réponse

Successful response

Response from the Responses API endpoint.

id
string
requis

Unique identifier for the response.

Exemple:

"resp_abc123"

object
enum<string>
requis

The object type.

Options disponibles:
response
created_at
integer
requis

Unix timestamp of when the response was created.

model
string
requis

The model used for the response.

status
enum<string>
requis

The status of the response.

Options disponibles:
completed,
failed,
in_progress,
cancelled,
incomplete
output
Reasoning Output · object · Message Output · object · Function Call Output · object · object · object · Web Search Call Output · object[]
requis

The output items generated by the model.

incomplete_details
object

Why generation ended before completion; partial output and usage are retained.

usage
Usage · object

Token usage statistics.

error
Error · object

Error information if the response failed.