Chat Completions

post https://api.perplexity.ai/chat/completions

Generates a model's response for the given chat conversation.

Body Params

model

string

required

Defaults to llama-3.1-sonar-small-128k-online

The name of the model that will complete your prompt. Possible values include llama-3.1-sonar-small-128k-chat, llama-3.1-sonar-small-128k-online, llama-3.1-sonar-large-128k-chat, llama-3.1-sonar-large-128k-online, llama-3.1-8b-instruct, and llama-3.1-70b-instruct.

messages

array of objects

required

Defaults to [object Object],[object Object]

A list of messages comprising the conversation so far.

Messages*

content

string

required

The contents of the message in this turn of conversation.

role

string

required

The role of the speaker in this turn of conversation. After the (optional) system message, user and assistant roles should alternate with user then assistant, ending in user.

content

string

required

The contents of the message in this turn of conversation.

role

string

required

The role of the speaker in this turn of conversation. After the (optional) system message, user and assistant roles should alternate with user then assistant, ending in user.

max_tokens

integer

The maximum number of completion tokens returned by the API. The total number of tokens requested in max_tokens plus the number of prompt tokens sent in messages must not exceed the context window token limit of model requested. If left unspecified, then the model will generate tokens until either it reaches its stop token or the end of its context window.

temperature

number

0 to 2

Defaults to 0.2

The amount of randomness in the response, valued between 0 inclusive and 2 exclusive. Higher values are more random, and lower values are more deterministic.

top_p

number

0 to 1

Defaults to 0.9

The nucleus sampling threshold, valued between 0 and 1 inclusive. For each subsequent token, the model considers the results of the tokens with top_p probability mass. We recommend either altering top_k or top_p, but not both.

return_citations

boolean

Defaults to false

Determines whether or not a request to an online model should return citations. Citations are in closed beta access. To request access, apply at https://perplexity.typeform.com/apiaccessform?typeform-source=docs.perplexity.ai

search_domain_filter

array

Defaults to null

Given a list of domains, limit the citations used by the online model to URLs from the specified domains. This feature is in closed beta access. To request access, apply at https://perplexity.typeform.com/apiaccessform?typeform-source=docs.perplexity.ai

Search Domain Filter

return_images

boolean

Defaults to false

Determines whether or not a request to an online model should return images. Images are in closed beta access. To request access, apply at https://perplexity.typeform.com/apiaccessform?typeform-source=docs.perplexity.ai

return_related_questions

boolean

Defaults to false

Determines whether or not a request to an online model should return related questions. Related questions are in closed beta access. To request access, apply at https://perplexity.typeform.com/apiaccessform?typeform-source=docs.perplexity.ai

top_k

number

0 to 2048

Defaults to 0

The number of tokens to keep for highest top-k filtering, specified as an integer between 0 and 2048 inclusive. If set to 0, top-k filtering is disabled. We recommend either altering top_k or top_p, but not both.

stream

boolean

Defaults to false

Determines whether or not to incrementally stream the response with server-sent events with content-type: text/event-stream.

presence_penalty

number

-2 to 2

Defaults to 0

A value 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. Incompatible with frequency_penalty.

frequency_penalty

number

> 0

Defaults to 1

A multiplicative penalty greater than 0. Values greater than 1.0 penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim. A value of 1.0 means no penalty. Incompatible with presence_penalty.

200OK

422Validation Error