FriendliAI Serverless.Completions API
The Serverless.Completions API from FriendliAI — 2 operation(s) for serverless.completions.
OpenAPI Specification
openapi: 3.1.0
info:
title: Friendli Suite API Reference Container.Audio Serverless.Completions API
description: This is an OpenAPI reference of Friendli Suite API.
termsOfService: https://friendli.ai/terms-of-service
contact:
name: FriendliAI Support Team
email: support@friendli.ai
version: 0.1.0
servers:
- url: https://api.friendli.ai
tags:
- name: Serverless.Completions
paths:
/serverless/v1/completions:
post:
tags:
- Serverless.Completions
summary: Completions
description: Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
operationId: serverlessCompletionsComplete
security:
- token: []
parameters:
- name: X-Friendli-Team
in: header
required: false
schema:
anyOf:
- type: string
- type: 'null'
description: ID of team to run requests as (optional parameter).
title: X-Friendli-Team
description: ID of team to run requests as (optional parameter).
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/ServerlessCompletionsBody'
responses:
'200':
description: Successfully generated a completion response.
content:
application/json:
schema:
$ref: '#/components/schemas/ServerlessCompletionsSuccess'
examples:
Example:
value:
id: cmpl-26a1e10db8544bc3adb488d2d205288b
model: zai-org/GLM-5.2
object: text_completion
choices:
- index: 0
seed: 42
text: This is indeed a test
tokens:
- 128000
- 2028
- 374
- 13118
- 264
- 1296
finish_reason: stop
usage:
prompt_tokens: 7
completion_tokens: 6
total_tokens: 13
'422':
description: Unprocessable Entity
x-speakeasy-name-override: complete
x-mint:
metadata:
title: Model APIs Completions
sidebarTitle: Completions
og:title: Model APIs Completions
description: Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
og:description: Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
href: /openapi/model-apis/completions
content: 'Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
See available models at [this pricing table](/guides/model-apis/pricing#billing-methods).
To request successfully, it is mandatory to enter a **Personal API Key** (e.g. flp_XXX) value in the **Bearer Token** field.
Refer to the [authentication section](/openapi/introduction#authentication) on our introduction page to learn how to acquire this variable and [visit here](https://friendli.ai/suite/~/setting/keys) to generate your API Key.
When streaming mode is used (i.e., `stream` option is set to `true`), the response is in MIME type `text/event-stream`. Otherwise, the content type is `application/json`.
You can view the schema of the streamed sequence of chunk objects in streaming mode [here](/openapi/model-apis/completions-chunk-object).
<Tip>You can explore examples on the [Friendli Model APIs](https://friendli.ai/get-started/model-apis) playground and adjust settings with just a few clicks.</Tip>'
/serverless/v1/completions#stream:
post:
tags:
- Serverless.Completions
summary: Stream completions
description: Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
operationId: serverlessCompletionsStream
security:
- token: []
parameters:
- name: X-Friendli-Team
in: header
required: false
schema:
anyOf:
- type: string
- type: 'null'
description: ID of team to run requests as (optional parameter).
title: X-Friendli-Team
description: ID of team to run requests as (optional parameter).
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/ServerlessCompletionsStreamBody'
responses:
'200':
description: Successfully generated a completion response.
content:
text/event-stream:
x-speakeasy-sse-sentinel: '[DONE]'
examples:
Example:
value: 'data: {"id":"cmpl-26a1e10db8544bc3adb488d2d205288b","object":"text_completion","choices":[{"index":0,"text":" such","token":1778,"finish_reason":null,"logprobs":null}],"created":1733382157}
data: {"id":"cmpl-26a1e10db8544bc3adb488d2d205288b","object":"text_completion","choices":[{"index":0,"text":" as","token":439,"finish_reason":null,"logprobs":null}],"created":1733382157}
...
data: {"id":"cmpl-26a1e10db8544bc3adb488d2d205288b","object":"text_completion","choices":[{"index":0,"text":"","finish_reason":"length","logprobs":null}],"created":1733382157}
data: {"id":"cmpl-26a1e10db8544bc3adb488d2d205288b","object":"text_completion","choices":[],"usage":{"prompt_tokens":5,"completion_tokens":10,"total_tokens":15},"created":1733382157}
data: [DONE]
'
schema:
$ref: '#/components/schemas/ServerlessCompletionsStreamSuccess'
'422':
description: Unprocessable Entity
x-speakeasy-name-override: stream
x-mint:
metadata:
title: Model APIs Stream Completions
sidebarTitle: Stream Completions
og:title: Model APIs Stream Completions
description: Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
og:description: Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
content: 'Generate a text completion from a prompt or token sequence. Supports streaming and configurable generation parameters.
See available models at [this pricing table](/guides/model-apis/pricing#billing-methods).
To request successfully, it is mandatory to enter a **Personal API Key** (e.g. flp_XXX) value in the **Bearer Token** field.
Refer to the [authentication section](/openapi/introduction#authentication) on our introduction page to learn how to acquire this variable and [visit here](https://friendli.ai/suite/~/setting/keys) to generate your API Key.
When streaming mode is used (i.e., `stream` option is set to `true`), the response is in MIME type `text/event-stream`. Otherwise, the content type is `application/json`.
You can view the schema of the streamed sequence of chunk objects in streaming mode [here](/openapi/model-apis/completions-chunk-object).
<Tip>You can explore examples on the [Friendli Model APIs](https://friendli.ai/get-started/model-apis) playground and adjust settings with just a few clicks.</Tip>'
components:
schemas:
CompletionsStreamServerlessBodyWithTokens:
properties:
model:
type: string
title: Model
description: Code of the model to use. See [available model list](https://friendli.ai/docs/guides/model-apis/pricing#billing-methods).
examples:
- zai-org/GLM-5.2
bad_word_tokens:
anyOf:
- items:
$ref: '#/components/schemas/TokenSequence'
type: array
- type: 'null'
title: Bad Word Tokens
description: Same as the above `bad_words` field, but receives token sequences instead of text phrases. This is similar to Hugging Face's [`bad_word_ids`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.bad_words_ids) argument.
bad_words:
anyOf:
- items:
type: string
type: array
- type: 'null'
title: Bad Words
description: 'Text phrases that should not be generated.
For a bad word phrase that contains N tokens, if the first N-1 tokens appears at the last of the generated result, the logit for the last token of the phrase is set to -inf.
Before checking whether a bard word is included in the result, the word is converted into tokens.
We recommend using `bad_word_tokens` because it is clearer.
For example, after tokenization, phrases "clear" and " clear" can result in different token sequences due to the prepended space character.
Defaults to empty list.
'
embedding_to_replace:
anyOf:
- items:
type: number
type: array
- type: 'null'
title: Embedding To Replace
description: A list of flattened embedding vectors used for replacing the tokens at the specified indices provided via `token_index_to_replace`.
encoder_no_repeat_ngram:
anyOf:
- type: integer
- type: 'null'
title: Encoder No Repeat Ngram
description: If this exceeds 1, every ngram of that size occurring in the input token sequence cannot appear in the generated result. 1 means that this mechanism is disabled (i.e., you cannot prevent 1-gram from being generated repeatedly). Only allowed for encoder-decoder models. Defaults to 1. This is similar to Hugging Face's [`encoder_no_repeat_ngram_size`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.encoder_no_repeat_ngram_size) argument.
encoder_repetition_penalty:
anyOf:
- type: number
- type: 'null'
title: Encoder Repetition Penalty
description: Penalizes tokens that have already appeared in the input tokens. Should be positive value. 1.0 means no penalty. Only allowed for encoder-decoder models. See [Keskar et al., 2019](https://arxiv.org/abs/1909.05858) for more details. This is similar to Hugging Face's [`encoder_repetition_penalty`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.encoder_repetition_penalty) argument.
eos_token:
anyOf:
- items:
type: integer
type: array
- type: 'null'
title: Eos Token
description: A list of endpoint sentence tokens.
forced_output_tokens:
anyOf:
- items:
type: integer
type: array
- type: 'null'
title: Forced Output Tokens
description: A token sequence that is enforced as a generation output. This option can be used when evaluating the model for the datasets with multi-choice problems (e.g., [HellaSwag](https://huggingface.co/datasets/hellaswag), [MMLU](https://huggingface.co/datasets/cais/mmlu)). Use this option with `logprobs` to get logprobs for the evaluation.
frequency_penalty:
anyOf:
- type: number
- type: 'null'
title: Frequency Penalty
description: Number between -2.0 and 2.0. Positive values penalizes tokens that have been sampled, taking into account their frequency in the preceding text. This penalization diminishes the model's tendency to reproduce identical lines verbatim.
logprobs:
anyOf:
- type: integer
- type: 'null'
title: Logprobs
description: Include the log probabilities on the logprobs most likely output tokens, as well the chosen tokens.
max_tokens:
anyOf:
- type: integer
- type: 'null'
title: Max Tokens
description: The maximum number of tokens to generate. For decoder-only models like GPT, the length of your input tokens plus `max_tokens` should not exceed the model's maximum length (e.g., 2048 for OpenAI GPT-3). For encoder-decoder models like T5 or BlenderBot, `max_tokens` should not exceed the model's maximum output length. This is similar to Hugging Face's [`max_new_tokens`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.max_new_tokens) argument.
max_total_tokens:
anyOf:
- type: integer
- type: 'null'
title: Max Total Tokens
description: The maximum number of tokens including both the generated result and the input tokens. Only allowed for decoder-only models. Only one argument between `max_tokens` and `max_total_tokens` is allowed. Default value is the model's maximum length. This is similar to Hugging Face's [`max_length`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.max_length) argument.
min_p:
anyOf:
- type: number
- type: 'null'
title: Min P
description: A scaling factor used to determine the minimum token probability threshold. This threshold is calculated as `min_p` multiplied by the probability of the most likely token. Tokens with probabilities below this scaled threshold are excluded from sampling. Values range from 0.0 (inclusive) to 1.0 (inclusive). Higher values result in stricter filtering, while lower values allow for greater diversity. The default value of 0.0 disables filtering, allowing all tokens to be considered for sampling.
min_tokens:
anyOf:
- type: integer
- type: 'null'
title: Min Tokens
description: 'The minimum number of tokens to generate. Default value is 0. This is similar to Hugging Face''s [`min_new_tokens`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.generationconfig.min_new_tokens) argument.
**This field is unsupported when `response_format` is specified.**'
min_total_tokens:
anyOf:
- type: integer
- type: 'null'
title: Min Total Tokens
description: The minimum number of tokens including both the generated result and the input tokens. Only allowed for decoder-only models. Only one argument between `min_tokens` and `min_total_tokens` is allowed. This is similar to Hugging Face's [`min_length`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.min_length) argument.
n:
anyOf:
- type: integer
- type: 'null'
title: N
description: The number of independently generated results for the prompt. Defaults to 1. This is similar to Hugging Face's [`num_return_sequences`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.num_return_sequences) argument.
no_repeat_ngram:
anyOf:
- type: integer
- type: 'null'
title: No Repeat Ngram
description: If this exceeds 1, every ngram of that size can only occur once among the generated result (plus the input tokens for decoder-only models). 1 means that this mechanism is disabled (i.e., you cannot prevent 1-gram from being generated repeatedly). Defaults to 1. This is similar to Hugging Face's [`no_repeat_ngram_size`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.no_repeat_ngram_size) argument.
presence_penalty:
anyOf:
- type: number
- type: 'null'
title: Presence Penalty
description: Number between -2.0 and 2.0. Positive values penalizes tokens that have been sampled at least once in the existing text.
repetition_penalty:
anyOf:
- type: number
- type: 'null'
title: Repetition Penalty
description: Penalizes tokens that have already appeared in the generated result (plus the input tokens for decoder-only models). Should be positive value (1.0 means no penalty). See [keskar et al., 2019](https://arxiv.org/abs/1909.05858) for more details. This is similar to Hugging Face's [`repetition_penalty`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.generationconfig.repetition_penalty) argument.
response_format:
anyOf:
- $ref: '#/components/schemas/ResponseFormat'
- type: 'null'
seed:
anyOf:
- items:
type: integer
type: array
- type: integer
- type: 'null'
title: Seed
description: Seed to control random procedure. If nothing is given, the API generate the seed randomly, use it for sampling, and return the seed along with the generated result. When using the `n` argument, you can pass a list of seed values to control all of the independent generations.
stop:
anyOf:
- items:
type: string
type: array
- type: 'null'
title: Stop
description: 'When one of the stop phrases appears in the generation result, the API will stop generation.
The stop phrases are excluded from the result.
Defaults to empty list.'
stop_tokens:
anyOf:
- items:
$ref: '#/components/schemas/TokenSequence'
type: array
- type: 'null'
title: Stop Tokens
description: Stop generating further tokens when generated token corresponds to any of the tokens in the sequence.
stream:
type: boolean
title: Stream
description: Whether to stream the generation result. When set to `true`, each token is sent as [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#event_stream_format) once generated.
default: true
stream_options:
anyOf:
- $ref: '#/components/schemas/StreamOptions'
- type: 'null'
description: 'Options related to stream.
It can only be used when `stream: true`.'
temperature:
anyOf:
- type: number
- type: 'null'
title: Temperature
description: Sampling temperature. Smaller temperature makes the generation result closer to greedy, argmax (i.e., `top_k = 1`) sampling. Defaults to 1.0. This is similar to Hugging Face's [`temperature`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.generationconfig.temperature) argument.
token_index_to_replace:
anyOf:
- items:
type: integer
type: array
- type: 'null'
title: Token Index To Replace
description: A list of token indices where to replace the embeddings of input tokens provided via either `tokens` or `prompt`.
top_k:
anyOf:
- type: integer
- type: 'null'
title: Top K
description: Limits sampling to the top k tokens with the highest probabilities. Values range from 0 (no filtering) to the model's vocabulary size (inclusive). The default value of 0 applies no filtering, allowing all tokens.
examples:
- 1
top_p:
anyOf:
- type: number
- type: 'null'
title: Top P
description: Keeps only the smallest set of tokens whose cumulative probabilities reach `top_p` or higher. Values range from 0.0 (exclusive) to 1.0 (inclusive). The default value of 1.0 includes all tokens, allowing maximum diversity.
xtc_threshold:
anyOf:
- type: number
- type: 'null'
title: Xtc Threshold
description: A probability threshold used to identify “top choice” tokens for exclusion in XTC (Exclude Top Choices) sampling. Tokens with probabilities at or above this threshold are considered viable candidates, and all but the least likely viable token are excluded from sampling. This option reduces the dominance of highly probable tokens while preserving some diversity by keeping the least confident “top choice.” Values range from 0.0 (inclusive) to 1.0 (inclusive). Higher values make the filtering more selective by requiring higher probabilities to trigger exclusion, while lower values apply filtering more broadly. The default value of 0.0 disables XTC filtering entirely.
xtc_probability:
anyOf:
- type: number
- type: 'null'
title: Xtc Probability
description: The probability that XTC (Exclude Top Choices) filtering will be applied for each sampling decision. When XTC is triggered, high-probability tokens above the `xtc_threshold` are excluded except for the least likely viable token. This stochastic activation allows for a balance between standard sampling and creativity-boosting exclusion filtering. Values range from 0.0 (inclusive) to 1.0 (inclusive), where 0.0 means XTC is never applied, 1.0 means XTC is always applied when viable tokens exist, and intermediate values provide probabilistic activation. The default value of 0.0 disables XTC filtering.
tokens:
items:
type: integer
type: array
title: Tokens
description: The tokenized prompt (i.e., input tokens). Either `prompt` or `tokens` field is required.
type: object
required:
- model
- tokens
title: CompletionsStreamServerlessBodyWithTokens
StreamedCompletionsResult:
properties:
data:
$ref: '#/components/schemas/StreamedCompletionData'
type: object
required:
- data
title: StreamedCompletionsResult
ServerlessCompletionsBody:
anyOf:
- $ref: '#/components/schemas/CompletionsServerlessBodyWithPrompt'
- $ref: '#/components/schemas/CompletionsServerlessBodyWithTokens'
title: ServerlessCompletionsBody
ServerlessCompletionsStreamSuccess:
$ref: '#/components/schemas/StreamedCompletionsResult'
title: ServerlessCompletionsStreamSuccess
StreamOptions:
properties:
include_usage:
anyOf:
- type: boolean
- type: 'null'
title: Include Usage
description: 'When set to `true`,
the number of tokens used will be included at the end of the stream result in the form of
`"usage": {"completion_tokens": number, "prompt_tokens": number, "total_tokens": number}`.
'
type: object
title: StreamOptions
ServerlessCompletionsSuccess:
$ref: '#/components/schemas/CompletionsResult'
title: ServerlessCompletionsSuccess
ResponseFormatJsonSchemaSchema:
properties:
schema:
additionalProperties: true
type: object
title: Schema
description: The schema for the response format, described as a JSON Schema object.
type: object
required:
- schema
title: ResponseFormatJsonSchemaSchema
PromptTokensDetails:
properties:
cached_tokens:
anyOf:
- type: integer
- type: 'null'
title: Cached Tokens
description: Cached tokens present in the prompt.
type: object
title: PromptTokensDetails
StreamedCompletionData:
properties:
id:
type: string
title: Id
description: A unique ID of the completion.
choices:
items:
$ref: '#/components/schemas/StreamedCompletionsChoice'
type: array
title: Choices
usage:
anyOf:
- $ref: '#/components/schemas/TextUsage'
- type: 'null'
object:
type: string
const: text_completion
title: Object
description: The object type, which is always set to `text_completion`.
created:
type: integer
title: Created
description: The Unix timestamp (in seconds) for when the token is sampled.
model:
anyOf:
- type: string
- type: 'null'
title: Model
description: The model to generate the completion. For dedicated endpoints, it returns the endpoint ID.
type: object
required:
- id
- choices
- object
- created
title: StreamedCompletionData
CompletionsResult:
properties:
id:
type: string
title: Id
description: A unique ID of the completion.
object:
type: string
const: text_completion
title: Object
description: The object type, which is always set to `text_completion`.
usage:
$ref: '#/components/schemas/TextUsage'
choices:
items:
$ref: '#/components/schemas/CompletionsChoice'
type: array
title: Choices
model:
anyOf:
- type: string
- type: 'null'
title: Model
description: The model to generate the completion. For dedicated endpoints, it returns the endpoint ID.
type: object
required:
- id
- object
- usage
- choices
title: CompletionsResult
TextUsage:
properties:
prompt_tokens:
type: integer
title: Prompt Tokens
description: Number of tokens in the prompt.
examples:
- 5
completion_tokens:
type: integer
title: Completion Tokens
description: Number of tokens in the generated completions.
examples:
- 7
total_tokens:
type: integer
title: Total Tokens
description: Total number of tokens used in the request (`prompt_tokens` + `completion_tokens`).
examples:
- 12
prompt_tokens_details:
anyOf:
- $ref: '#/components/schemas/PromptTokensDetails'
- type: 'null'
description: Breakdown of tokens used in the prompt.
type: object
required:
- prompt_tokens
- completion_tokens
- total_tokens
title: TextUsage
ResponseFormat:
oneOf:
- $ref: '#/components/schemas/ResponseFormatJsonSchema'
title: Json Schema
- $ref: '#/components/schemas/ResponseFormatJsonObject'
title: Json Object
- $ref: '#/components/schemas/ResponseFormatRegex'
title: Regex
- $ref: '#/components/schemas/ResponseFormatText'
title: Text
description: 'The enforced format of the model''s output.
Note that the content of the output message may be truncated if it exceeds the `max_tokens`. You can check this by verifying that the `finish_reason` of the output message is `length`.
For more detailed information, please refer [here](https://friendli.ai/docs/guides/structured-outputs).
***Important***
You must explicitly instruct the model to produce the desired output format using a system prompt or user message (e.g., `You are an API generating a valid JSON as output.`).
Otherwise, the model may result in an unending stream of whitespace or other characters.
**When `response_format` is specified, `min_tokens` field is unsupported.**'
discriminator:
propertyName: type
mapping:
json_object: '#/components/schemas/ResponseFormatJsonObject'
json_schema: '#/components/schemas/ResponseFormatJsonSchema'
regex: '#/components/schemas/ResponseFormatRegex'
text: '#/components/schemas/ResponseFormatText'
CompletionsStreamServerlessBodyWithPrompt:
properties:
model:
type: string
title: Model
description: Code of the model to use. See [available model list](https://friendli.ai/docs/guides/model-apis/pricing#billing-methods).
examples:
- zai-org/GLM-5.2
bad_word_tokens:
anyOf:
- items:
$ref: '#/components/schemas/TokenSequence'
type: array
- type: 'null'
title: Bad Word Tokens
description: Same as the above `bad_words` field, but receives token sequences instead of text phrases. This is similar to Hugging Face's [`bad_word_ids`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.bad_words_ids) argument.
bad_words:
anyOf:
- items:
type: string
type: array
- type: 'null'
title: Bad Words
description: 'Text phrases that should not be generated.
For a bad word phrase that contains N tokens, if the first N-1 tokens appears at the last of the generated result, the logit for the last token of the phrase is set to -inf.
Before checking whether a bard word is included in the result, the word is converted into tokens.
We recommend using `bad_word_tokens` because it is clearer.
For example, after tokenization, phrases "clear" and " clear" can result in different token sequences due to the prepended space character.
Defaults to empty list.
'
embedding_to_replace:
anyOf:
- items:
type: number
type: array
- type: 'null'
title: Embedding To Replace
description: A list of flattened embedding vectors used for replacing the tokens at the specified indices provided via `token_index_to_replace`.
encoder_no_repeat_ngram:
anyOf:
- type: integer
- type: 'null'
title: Encoder No Repeat Ngram
description: If this exceeds 1, every ngram of that size occurring in the input token sequence cannot appear in the generated result. 1 means that this mechanism is disabled (i.e., you cannot prevent 1-gram from being generated repeatedly). Only allowed for encoder-decoder models. Defaults to 1. This is similar to Hugging Face's [`encoder_no_repeat_ngram_size`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.encoder_no_repeat_ngram_size) argument.
encoder_repetition_penalty:
anyOf:
- type: number
- type: 'null'
title: Encoder Repetition Penalty
description: Penalizes tokens that have already appeared in the input tokens. Should be positive value. 1.0 means no penalty. Only allowed for encoder-decoder models. See [Keskar et al., 2019](https://arxiv.org/abs/1909.05858) for more details. This is similar to Hugging Face's [`encoder_repetition_penalty`](https://huggingface.co/docs/transformers/v4.26.0/en/main_classes/text_generation#transformers.GenerationConfig.encoder_repetition_penalty) argument.
eos_token:
anyOf:
- items:
type: integer
type: array
- type: 'null'
title: Eos Token
description: A list of endpoint sentence tokens.
forced_output_tokens:
anyOf:
- items:
type: integer
type: array
- type: 'null'
title: Forced Output Tokens
description: A token sequence that is enforced as a generation output. This option can be used when evaluating the model for the datasets with multi-choice problems (e.g., [HellaSwag](https://huggingface.co/datasets/hellaswag), [MMLU](https://huggingface.co/datasets/cais/mmlu)). Use this option with `logprobs` to get logprobs for the evaluation.
frequency_penalty:
anyOf:
- type: number
- type: 'null'
title: Frequency Penalty
description: Number betwe
# --- truncated at 32 KB (77 KB total) ---
# Full source: https://raw.githubusercontent.com/api-evangelist/friendliai/refs/heads/main/openapi/friendliai-serverless-completions-api-openapi.yml