Chat Completions

post

https://api.konko.ai/v1/chat/completions

Use this endpoint with chat models tailored to generate the next message in a conversation.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Body Params

model

enum

required

The chat model used to generate a completion.

Konko currently offers open source LLMs hosted by Konko and recommended LLMs provided by OpenAI.

See Available Models for details on which models work best for your use case.

messages

array of objects

required

length ≥ 1

A list of messages the chat model uses to generate the next completion.

See our prompting guidelines for more information.

Use the "Add" button below to add more messages. For each message added, you select one type of role for that message.

A typical conversation would start with a system and user message, then have repeating assistant and user messages.

messages*

content

string | null

required

Defaults to Please summarize the Harry Potter series for me.

The contents of the user message.

role

string

enum

required

The role of the messages author, in this case user.

Allowed:

name

string

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

max_tokens

integer | null

Defaults to 24

The chat completion's token limit, combined with input tokens, is bound by the model's context length. Different models have varying limits, with a common maximum being 2048 tokens, unless otherwise specified.

stream

boolean | null

Defaults to false

When stream is set to True, the API sends partial message updates via HTTP streaming as they're available, useful for real-time data or large messages benefiting from incremental delivery.

temperature

number | null

0 to 2

Defaults to 1

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_p

number | null

0 to 1

Defaults to 1

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.

frequency_penalty

number | null

-2 to 2

Defaults to 0

Currently, only OpenAI provided models support frequency_penalty. Use repetition_penalty with all other models.

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

repetition_penalty

number | null

-2 to 2

Defaults to 0

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

logit_bias

object | null

Currently, only OpenAI provided models support logit_bias.

Adjusts the chances of specific tokens in the completion.

Uses a JSON object mapping tokens (by their ID) to a bias value (-100 to 100).

This bias tweaks the model's output before sampling.

Values between -1 and 1 subtly change token likelihood.

Values near -100 or 100 typically ban or exclusively pick that token.

integer | null

1 to 128

Defaults to 1

How many chat completion choices to generate for each input message.
n currently only works for OpenAI provided models.
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.

presence_penalty

number | null

-2 to 2

Defaults to 0

Currently, only OpenAI provided models support presence_penalty. Use repetition_penalty with all other models.

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.

seed

integer | null

-9223372036854776000 to 9223372036854776000

Currently, only OpenAI provided models support seed.

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.

stop

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

200OK