Create a chat completion

post

https://api.cynos.averlon.ai/v1/chat/completions

Creates a completion for the provided messages. Supports streaming via SSE.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Body Params

model

string

required

Public model id. An OpenAI model is openai/, for example openai/gpt-4o.

messages

array of objects

required

List of messages in the conversation

messages*

stream

boolean

Whether to stream the response

frequency_penalty

number

-2 to 2

logit_bias

object

logprobs

boolean

max_completion_tokens

integer

metadata

object

modalities

array of strings

modalities

parallel_tool_calls

boolean

presence_penalty

number

-2 to 2

prompt_cache_key

string

Prompt cache key. OpenAI provider only. When set, routing keeps the provider on OpenAI and may still change the model.

reasoning_effort

string

enum

Alias of reasoning.effort. Cannot be sent together with reasoning.effort.

Allowed:

response_format

object

Format for the response

safety_identifier

string

service_tier

string

stream_options

object

store

boolean

Whether to store the completion. OpenAI provider only. When set, routing keeps the provider on OpenAI and may still change the model.

temperature

number

0 to 2

tool_choice

enum

tools

array of objects

tools

seed

integer

Deterministic sampling seed

top_p

number

0 to 1

Nucleus sampling parameter

top_logprobs

integer

0 to 20

Number of most likely tokens to return at each position

stop

Up to 4 sequences where the API will stop generating tokens

prediction

object

Predicted output content for the model to reference. OpenAI provider only. When set, routing keeps the provider on OpenAI and may still change the model. Can reduce latency.

prompt_cache_retention

string

enum

Prompt cache retention policy. OpenAI provider only. When set, routing keeps the provider on OpenAI and may still change the model.

Allowed:

prompt_cache_options

object

Request-wide prompt cache options (OpenAI gpt-5.6+). OpenAI provider only. When set, routing keeps the provider on OpenAI and may still change the model.

web_search_options

object

Web search options for chat completions. Supported on the OpenAI and Gemini providers. The route is limited to a model that can serve the search.

user

string

verbosity

string

enum

Response verbosity. OpenAI provider only. When set, routing keeps the provider on OpenAI and may still change the model.

Allowed:

audio

object

Audio output parameters. Supported by OpenAI models that accept audio output.

integer

Number of chat completions to generate when the provider supports it

Headers

x-cy-eh-quality-judge

string

enum

Force the quality judge on or off for this request. off skips evaluation and ignores judge_effort. Any value other than on or off is ignored and the plugin's enable switch is used.

Allowed:

x-cy-eh-task-route

string

enum

Force task routing on or off for this request. on routes even when the plugin default is off. off uses the caller's model. Any value other than on or off is ignored.

Allowed:

string

enum

Defaults to application/json

Generated from available response content types

Allowed:

Responses

200Successful response

400Bad request

500Internal server error