Generate chat-style completions from open-source LLMs
openapi: 3.1.0
info:
title: Hyperbolic Audio Generation Chat Completions API
description: 'Convert text to natural-sounding speech using audio models hosted by Hyperbolic — Melo TTS (sunset) and Whisper (coming soon). Returns base64-encoded audio. Pricing from $0.001 per 1000 characters.
'
version: v1
contact:
name: Hyperbolic Support
email: support@hyperbolic.ai
url: https://docs.hyperbolic.ai
license:
name: Hyperbolic Terms of Use
url: https://www.hyperbolic.ai/terms-of-use
servers:
- url: https://api.hyperbolic.xyz/v1
description: Hyperbolic Production Inference Server
security:
- BearerAuth: []
tags:
- name: Chat Completions
description: Generate chat-style completions from open-source LLMs
paths:
/chat/completions:
post:
summary: Hyperbolic Create A Chat Completion
description: 'Create a chat completion using one of the open-source LLMs served by Hyperbolic. The request and response shape is fully OpenAI-compatible: send a list of messages with role (`system`, `user`, `assistant`, `tool`) and content (text or, for vision models, image_url blocks). Streaming is enabled by setting `stream: true`. Tool calling, structured outputs (`response_format`), and reasoning models (DeepSeek R1) are supported.
'
operationId: createChatCompletion
tags:
- Chat Completions
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/ChatCompletionRequest'
examples:
SimpleChat:
summary: Simple chat with DeepSeek V3
value:
model: deepseek-ai/DeepSeek-V3
messages:
- role: user
content: What is the capital of France?
max_tokens: 256
temperature: 0.7
Llama405B:
summary: Llama 3.1 405B reasoning
value:
model: meta-llama/Meta-Llama-3.1-405B-Instruct
messages:
- role: system
content: You are a careful research assistant.
- role: user
content: Summarize the difference between BF16 and FP8 inference.
max_tokens: 512
temperature: 0.5
stream: false
ToolUse:
summary: Function calling
value:
model: meta-llama/Meta-Llama-3.1-70B-Instruct
messages:
- role: user
content: What is the weather in Reston, VA?
tools:
- type: function
function:
name: get_weather
description: Get current weather for a location
parameters:
type: object
properties:
location:
type: string
description: City and state
required:
- location
tool_choice: auto
responses:
'200':
description: Successful chat completion response
content:
application/json:
schema:
$ref: '#/components/schemas/ChatCompletionResponse'
examples:
SimpleResponse:
summary: Simple chat completion
value:
id: chatcmpl-abc123
object: chat.completion
created: 1748133600
model: deepseek-ai/DeepSeek-V3
choices:
- index: 0
message:
role: assistant
content: The capital of France is Paris.
finish_reason: stop
usage:
prompt_tokens: 12
completion_tokens: 7
total_tokens: 19
text/event-stream:
schema:
type: string
description: 'Server-sent events stream when `stream: true`. Each event is a JSON-encoded ChatCompletionChunk with a `choices[].delta` payload, terminated by `data: [DONE]`.
'
'400':
description: Bad Request — invalid model, parameters, or message structure
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
'401':
description: Unauthorized — missing or invalid Bearer token
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
'402':
description: Payment Required — account balance exhausted
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
'429':
description: Too Many Requests — rate limit exceeded
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
'500':
description: Internal Server Error
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
components:
schemas:
ChatCompletionChoice:
type: object
properties:
index:
type: integer
message:
$ref: '#/components/schemas/ChatMessage'
finish_reason:
type: string
enum:
- stop
- length
- tool_calls
- content_filter
ChatCompletionResponse:
type: object
required:
- id
- object
- created
- model
- choices
properties:
id:
type: string
object:
type: string
enum:
- chat.completion
created:
type: integer
description: Unix epoch seconds.
model:
type: string
choices:
type: array
items:
$ref: '#/components/schemas/ChatCompletionChoice'
usage:
$ref: '#/components/schemas/Usage'
system_fingerprint:
type: string
Tool:
type: object
required:
- type
- function
properties:
type:
type: string
enum:
- function
function:
type: object
required:
- name
properties:
name:
type: string
description:
type: string
parameters:
type: object
description: JSON Schema describing the function arguments.
ChatCompletionRequest:
type: object
required:
- model
- messages
properties:
model:
type: string
description: 'Hyperbolic model identifier — see GET /v1/models for the live catalog. Examples: `meta-llama/Meta-Llama-3.1-405B-Instruct`, `meta-llama/Meta-Llama-3.1-70B-Instruct`, `deepseek-ai/DeepSeek-V3`, `deepseek-ai/DeepSeek-R1`, `Qwen/Qwen2.5-72B-Instruct`, `NousResearch/Hermes-3-Llama-3.1-70B`, `mistralai/Mistral-7B-Instruct`, `meta-llama/Llama-3.2-90B-Vision-Instruct`.
'
messages:
type: array
minItems: 1
items:
$ref: '#/components/schemas/ChatMessage'
max_tokens:
type: integer
minimum: 1
description: Maximum number of tokens to generate. Defaults vary by model.
temperature:
type: number
minimum: 0
maximum: 2
default: 1
top_p:
type: number
minimum: 0
maximum: 1
default: 1
top_k:
type: integer
minimum: 0
n:
type: integer
minimum: 1
default: 1
description: Number of completions to generate.
stream:
type: boolean
default: false
description: Stream partial deltas as Server-Sent Events.
stop:
oneOf:
- type: string
- type: array
items:
type: string
description: Stop sequence(s) — up to 4.
presence_penalty:
type: number
minimum: -2
maximum: 2
frequency_penalty:
type: number
minimum: -2
maximum: 2
seed:
type: integer
description: Best-effort deterministic seed.
response_format:
type: object
description: 'Structured-output specification (e.g. { "type": "json_object" }).'
properties:
type:
type: string
enum:
- text
- json_object
- json_schema
tools:
type: array
items:
$ref: '#/components/schemas/Tool'
description: Function/tool catalog for OpenAI-compatible tool calling.
tool_choice:
oneOf:
- type: string
enum:
- none
- auto
- required
- type: object
user:
type: string
description: End-user identifier for abuse monitoring.
Usage:
type: object
properties:
prompt_tokens:
type: integer
completion_tokens:
type: integer
total_tokens:
type: integer
ToolCall:
type: object
required:
- id
- type
- function
properties:
id:
type: string
type:
type: string
enum:
- function
function:
type: object
properties:
name:
type: string
arguments:
type: string
description: JSON-encoded argument string.
ErrorResponse:
type: object
properties:
error:
type: object
properties:
message:
type: string
type:
type: string
code:
type: string
param:
type: string
ChatMessage:
type: object
required:
- role
- content
properties:
role:
type: string
enum:
- system
- user
- assistant
- tool
content:
oneOf:
- type: string
- type: array
items:
$ref: '#/components/schemas/ContentPart'
name:
type: string
tool_call_id:
type: string
tool_calls:
type: array
items:
$ref: '#/components/schemas/ToolCall'
ContentPart:
type: object
required:
- type
properties:
type:
type: string
enum:
- text
- image_url
text:
type: string
image_url:
type: object
properties:
url:
type: string
description: Public URL or `data:` URI for image input on vision models.
detail:
type: string
enum:
- auto
- low
- high
securitySchemes:
BearerAuth:
type: http
scheme: bearer
bearerFormat: API Key