Chat Completions - Cerebras Inference

Authorizations

Authorization

string

header

required

API key for authentication. Obtain your key from the Cerebras Cloud console and pass it as Authorization: Bearer YOUR_API_KEY.

Headers

Content-Type

enum<string>

default:application/json

The media type of the request body. Supported values: application/json, application/vnd.msgpack. Default: application/json.

Available options:

application/json,

application/vnd.msgpack

Content-Encoding

enum<string>

The compression encoding applied to the request body. When set, the request body must be gzip-compressed. Can be combined with any supported Content-Type. Supported values: gzip.

Available options:

gzip

queue_threshold

string

Controls the queue time threshold for requests using the flex or auto service tiers. Requests are preemptively rejected if the rolling average queue time exceeds this threshold. Valid range: 50–20000 (milliseconds). Private Preview.

Body

application/json

messages

(System message · object | User message · object | Assistant message · object | Tool message · object | Developer message · object)[]

required

A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text and images.

Developer-provided instructions the model should follow regardless of user messages.

System message
User message
Assistant message
Tool message
Developer message

Hide child attributes

messages.role

enum<string>

required

The role of the message author. Always system.

Available options:

system

messages.content

required

The contents of the system message.

messages.name

string

An optional name for the participant. Provides the model with information to differentiate between participants of the same role.

model

string

required

The ID of the model to use for generating a response.

clear_thinking

boolean | null

Controls whether thinking content from previous conversation turns is included in the prompt context.

false - Thinking from all previous turns is preserved in the conversation history. Recommended for agentic workflows where reasoning from past tool-calling turns may be relevant for future tool calls.
true (default) - Thinking from earlier turns is excluded. Recommended for general chat conversations where reasoning from past turns is less relevant for performance.

When this parameter is not specified or set to null, the API defaults to clear_thinking: true.

Only supported on the zai-glm-4.7 model. For additional information, see Preserved thinking in the Z.ai documentation.

frequency_penalty

number | null

default:0

A number between -2.0 and 2.0. Positive values reduce the likelihood of the model repeating tokens by applying a penalty proportional to how frequently each token has already appeared in the generated output.

Required range: -2 <= x <= 2

logit_bias

object

Modify the likelihood of specified tokens appearing in the completion. Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.

Hide child attributes

logit_bias.{key}

number

logprobs

boolean

default:false

Whether to return log probabilities of the output tokens.

max_completion_tokens

integer | null

The maximum number of tokens that can be generated in the completion, including reasoning tokens. The total length of input tokens and generated tokens is limited by the model's context length.

parallel_tool_calls

boolean | null

default:true

Whether to enable parallel function calling during tool use. When enabled (default), the model can request multiple tool calls simultaneously in a single response. When disabled, only one tool call is made at a time.

prediction

object

Configuration for a Predicted Output, which can greatly speed up response times when large parts of the model response are known in advance. Most common when regenerating a file with mostly minor changes.

Hide child attributes

prediction.type

enum<string>

required

The type of predicted content. Always content.

Available options:

content

prediction.content

required

The content that should be matched when generating a model response.

presence_penalty

number | null

default:0

A number between -2.0 and 2.0. Positive values reduce the likelihood of the model repeating tokens that have already appeared in the output, encouraging the model to introduce new topics.

Required range: -2 <= x <= 2

prompt_cache_key

string | null

An opaque identifier that groups related requests so they reuse the same prompt cache. Requests sharing the same 'prompt_cache_key' are routed together, which increases cache hits and reduces time to first token.

Set it to a stable identifier like a conversation ID, user ID, or session ID.

Requires account-level enablement. Contact us or reach out to your account representative to request access.

Maximum string length: 1024

reasoning_effort

enum<string> | null

Controls the amount of reasoning the model performs. Supported values vary by model:

gpt-oss-120b: low, medium (default), high
zai-glm-4.7: none (disables reasoning)
gemma-4-31b: none (default), low, medium, high (all enable reasoning equivalently). raw and hidden reasoning formats are not supported.

Available options:

low,

medium,

high,

none

response_format

Text · object

An object that controls the format of the model response.

Setting { "type": "json_schema", ... } enables Structured Outputs, which enforces schema compliance. See Structured Outputs for details.

Setting { "type": "json_object" } enables legacy JSON mode, which ensures the model returns valid JSON but does not enforce a specific schema. To use json_object, include a system or user message specifying the desired format. json_object is not compatible with streaming — stream must be set to false.

Text
JSON Schema
JSON Object

Hide child attributes

response_format.type

enum<string>

required

The type of response format being defined. Always text.

Available options:

text

seed

integer | null

If specified, the system will make a best effort to sample deterministically so repeated requests with the same seed and parameters return the same result. Determinism is not guaranteed.

service_tier

enum<string> | null

default:default

Controls request prioritization.

Note: This feature is in Private Preview. For access or more information, contact us or reach out to your account representative.

Available options:

priority - Highest priority processing (Only available for dedicated endpoints, not shared endpoints.)
default - Standard priority processing
auto - Automatically uses the highest available service tier
flex - Lowest priority processing

See Service Tiers for more information.

Available options:

priority,

default,

auto,

flex

stop

string | null

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stream

boolean | null

If set to true, partial message deltas will be sent.

temperature

number | null

Sampling temperature between 0 and 2.0. Higher values (e.g. 0.8) make output more random; lower values (e.g. 0.2) make it more focused and deterministic. We recommend altering this or top_p, not both.

Required range: 0 <= x <= 2

tool_choice

Controls which (if any) tool is called by the model.

none — no tool is called
auto — (default when tools are present) the model chooses whether to call a tool
required — forces a tool call

A specific tool can be forced by passing an object specifying the function name.

Available options:

none,

auto,

required

tools

object[] | null

A list of tools the model may call. Use this to provide a list of functions the model may generate JSON inputs for.

Currently, only functions are supported. Specifying tools consumes prompt tokens; too many may degrade performance or hit context length limits.

Hide child attributes

tools.type

enum<string>

required

The type of the tool. Currently only function is supported.

Available options:

function

tools.function

object

required

Hide child attributes

tools.function.name

string

required

The name of the function to call. Supported characters: a-z, A-Z, 0-9, _, -. Maximum length: 64.

tools.function.description

string

A description of what the function does, used by the model to choose when and how to call the function.

tools.function.parameters

object

The parameters the function accepts, described as a JSON Schema object. Omitting parameters defines a function with an empty parameter list.

top_logprobs

integer | null

An integer between 0 and 20 specifying the number of most likely tokens to return at each token position, each with an associated log probability. If using this parameter, logprobs must also be set to true.

Required range: 0 <= x <= 20

top_p

number | null

Nucleus sampling parameter. The model considers only the tokens comprising the top top_p probability mass (e.g. 0.1 means only the tokens comprising the top 10% probability mass are considered). We recommend altering this or temperature, not both.

Required range: 0 <= x <= 1

user

string | null

A unique identifier representing your end-user, which can help monitor and detect abuse.

Response

200 - application/json

Successful response

string

A unique identifier for the chat completion.

choices

object[]

A list of chat completion choices. Can be more than one if n is greater than 1.

Hide child attributes

choices.finish_reason

enum<string>

The reason the model stopped generating tokens.

Available options:

stop,

length,

content_filter,

tool_calls

choices.index

integer

The index of the choice in the list of choices.

choices.logprobs

object

Log probability information for the output tokens.

choices.reasoning_logprobs

object

Log probability information for the reasoning tokens, if provided by the model.

choices.message

object

A chat completion message generated by the model.

Hide child attributes

choices.message.content

string | null

The contents of the message.

choices.message.role

string

The role of the author of this message.

choices.message.reasoning

string

The model's reasoning content when using reasoning models.

choices.message.tool_calls

object[] | null

Tool calls the model requested in its response.

Hide child attributes

choices.message.tool_calls.id

string

The ID of the tool call.

choices.message.tool_calls.type

enum<string>

The type of tool call.

Available options:

function

choices.message.tool_calls.function

object

The function call details.

Hide child attributes

choices.message.tool_calls.function.name

string

The name of the function called.

choices.message.tool_calls.function.arguments

string

The JSON-encoded arguments supplied by the model.

created

integer

Unix timestamp (in seconds) of when the completion was created.

model

string

The model used for the chat completion.

object

enum<string>

The object type. Always chat.completion.

Available options:

chat.completion

system_fingerprint

string

A fingerprint for the model or backend used to generate the response.

service_tier

string | null

The service tier used for the request, or null if not specified.

service_tier_used

enum<string> | null

The service tier used for processing the request. Only present when service_tier is set to auto in the request.

Available options:

priority,

default,

flex

usage

object

Usage statistics for the completion request.

Hide child attributes

usage.prompt_tokens

integer

Number of tokens in the prompt.

usage.completion_tokens

integer

Number of tokens in the generated completion.

usage.total_tokens

integer

Total number of tokens used (prompt + completion).

usage.image_tokens

integer

Number of tokens used to represent image inputs. Present for requests to vision-capable models.

usage.prompt_tokens_details

object

Detailed breakdown of prompt token usage.

Hide child attributes

usage.prompt_tokens_details.cached_tokens

integer

Number of prompt tokens served from the cache, reused from a previous request.

usage.completion_tokens_details

object

Breakdown of completion tokens when using Predicted Outputs.

Hide child attributes

usage.completion_tokens_details.accepted_prediction_tokens

integer

Tokens in the prediction that appeared in the completion.

usage.completion_tokens_details.rejected_prediction_tokens

integer

Tokens in the prediction that did not appear in the completion. Counted in total completion tokens for billing, output, and context window limits.

usage.completion_tokens_details.reasoning_tokens

integer

Tokens spent on model reasoning, if applicable.

time_info

object

Performance timing information for the request.

Hide child attributes

time_info.queue_time

number

Time spent in queue waiting for processing (seconds).

time_info.prompt_time

number

Time spent processing prompt/input tokens (seconds).

time_info.completion_time

number

Time spent generating completion/output tokens (seconds).

time_info.total_time

number

Total time for the entire request from submission to completion (seconds).

time_info.created

number

Unix timestamp of when the time_info was recorded.