v51

latestOpenAPI 3.1.0raw.githubusercontent.com2026-07-2177185.9 KB

Chat

post/v1/chat/completions

Headers

X-Amz-Cf-Idstring nullable

CF-RAYstring nullable

X-delay-timenumber nullable

Request body

clear_thinkingboolean nullable

When True, removes reasoning content from messages that appear before the latest user message.

disable_reasoningboolean nullable

Disables reasoning for reasoning models. If set to True, the model will not use any reasoning in its response.

frequency_penaltynumber nullable

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

logit_biasobject nullable

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.

logprobsboolean nullable

Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.

max_completion_tokensinteger nullable

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensinteger nullable

The maximum number of tokens that can be generated in the chat completion. The total length of input tokens and generated tokens is limited by the model's context length. This value is now deprecated in favor of max_completion_tokens.

min_completion_tokensinteger nullable

The minimum number of tokens to generate for a completion. If not specified or set to 0, the model will generate as many tokens as it deems necessary. Setting to -1 sets to max sequence length.

min_tokensinteger nullable

The minimum number of tokens to generate for a completion. If not specified or set to 0, the model will generate as many tokens as it deems necessary. Setting to -1 sets to max sequence length.

modelstring required

model_parametersobject nullable

Model-specific parameters. The accepted keys and values are defined per-model by the model's model_parameters schema. The schema is used downstream to validate the contents of this field. Unsupported for models that declare none.

ninteger nullable

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

parallel_tool_callsboolean nullable

presence_penaltynumber nullable

Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

prompt_cache_keystring nullable

An optional opaque string. The requests with the same prompt cache key would highly likely share the same prompt prefixes. Examples would be IDs of chat conversations, IDs of users, the hashes of system prompts, etc.

reasoning_effort'none' | 'low' | 'medium' | 'high'

reasoning_format'none' | 'parsed' | 'text_parsed' | 'raw' | 'hidden'

seedinteger nullable

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

service_tier'auto' | 'default' | 'flex' | 'priority'

streamboolean nullable

temperaturenumber nullable

What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or top_p but not both.

top_logprobsinteger nullable

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. logprobs must be set to true if this parameter is used.

top_pnumber nullable

An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or temperature but not both.

userstring nullable

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

Example request

{
  "max_completion_tokens": -1,
  "messages": [
    {
      "content": "You are a helpful assistant running on a CS-3 hardware at Cerebras Systems",
      "role": "system"
    },
    {
      "content": "What is Generative AI?",
      "role": "user"
    }
  ],
  "min_tokens": 1000,
  "model": "llama3.1-8b",
  "seed": 0,
  "stop_sequence": null,
  "stream": false,
  "temperature": 0,
  "top_p": 1
}

Response

Successful Response

Example response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "message": {
        "content": "Generative AI is a subfield of artificial intelligence that focuses on generating new, original content, such as images, music, text, or videos. These models are trained on large datasets and learn to recognize patterns, relationships, and structures within the data, allowing them to generate new content that is similar in style, tone, or characteristics to the training data. \n \n Generative AI models can be categorized into two main types: Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs). GANs consist of two neural networks: a generator and a discriminator. The generator creates new content, while the discriminator evaluates the generated content and tells the generator whether it's realistic or not. VAEs are neural networks that learn to compress and reconstruct data. \n \n Generative AI has numerous applications across various industries, including art and design, content generation, data augmentation, and recommendation systems. Some examples of Generative AI in action include Deep Dream Generator, Amper Music, and DALL-E. \n \n In summary, Generative AI is a powerful technology that enables the creation of new, original content using machine learning models. It has the potential to revolutionize various industries and create new opportunities for artists, designers, and content creators.",
        "role": "assistant"
      }
    }
  ],
  "created": 1721088719,
  "id": "response_1721",
  "model": "llama3.1-8b",
  "object": "chat.completion",
  "system_fingerprint": "88719",
  "time_info": {
    "completion_time": 0.44,
    "prompt_time": 0.02,
    "queue_time": 0,
    "total_time": 0.46
  },
  "usage": {
    "completion_tokens": 1020,
    "prompt_tokens": 40,
    "prompt_tokens_details": {
      "cached_tokens": 1020
    },
    "total_tokens": 1060
  }
}