Image generation (diffusion), embeddings, audio, and other non-chat model workloads are exposed as user-defined "cords" (decorated HTTP endpoints) on each deployed chute, plus templated vLLM and diffusion deploy helpers.
openapi: 3.0.1
info:
title: Chutes API
description: >-
Specification of the Chutes API. Chutes is a permissionless, serverless AI
compute platform on Bittensor Subnet 64. It exposes two REST surfaces: an
OpenAI-compatible inference endpoint at https://llm.chutes.ai/v1 for chat
completions and model listing across hundreds of open-source models, and a
management API at https://api.chutes.ai for building images and deploying,
listing, and operating chutes (serverless AI apps). All requests authenticate
with a Bearer API key (prefix `cpk_`).
termsOfService: https://chutes.ai/terms
contact:
name: Chutes Support
url: https://chutes.ai/docs
version: '1.0'
servers:
- url: https://llm.chutes.ai/v1
description: OpenAI-compatible LLM inference endpoint (chat completions, models).
- url: https://api.chutes.ai
description: Management API for chutes and images.
security:
- bearerAuth: []
tags:
- name: Chat
description: OpenAI-compatible chat completions.
- name: Models
description: List models available on the Chutes network.
- name: Chutes
description: Deploy, list, retrieve, update, and delete chutes.
- name: Images
description: Build, list, retrieve, and delete container images.
paths:
/chat/completions:
post:
operationId: createChatCompletion
servers:
- url: https://llm.chutes.ai/v1
tags:
- Chat
summary: Create a chat completion.
description: >-
OpenAI-compatible chat completions. Reachable for any model currently
running on the Chutes network via the shared endpoint
https://llm.chutes.ai/v1. Set `stream: true` to receive the response as
Server-Sent Events.
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/CreateChatCompletionRequest'
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/CreateChatCompletionResponse'
text/event-stream:
schema:
type: string
description: >-
Server-Sent Events stream of `chat.completion.chunk` objects
when `stream` is true, terminated by `data: [DONE]`.
'401':
description: Unauthorized - missing or invalid API key.
'422':
description: Validation error.
'429':
description: Too Many Requests - rate or quota limit exceeded.
/models:
get:
operationId: listModels
servers:
- url: https://llm.chutes.ai/v1
tags:
- Models
summary: List available models.
description: Lists the live models currently reachable on the Chutes network.
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/ListModelsResponse'
/chutes/:
get:
operationId: listChutes
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: List chutes.
description: List (and optionally filter/paginate) chutes.
parameters:
- name: include_public
in: query
required: false
schema:
type: boolean
- name: template
in: query
required: false
schema:
type: string
- name: name
in: query
required: false
schema:
type: string
- name: image
in: query
required: false
schema:
type: string
- name: slug
in: query
required: false
schema:
type: string
- name: page
in: query
required: false
schema:
type: integer
- name: limit
in: query
required: false
schema:
type: integer
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/ChuteList'
'422':
description: Validation error.
post:
operationId: deployChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Deploy a chute.
description: >-
Standard deploy from the Chutes development kit (CDK). Stores the chute
code, image reference, node selector, and configuration, then schedules
it onto suitable miner GPU capacity.
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/DeployChuteRequest'
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/Chute'
'422':
description: Validation error.
/chutes/{chute_id_or_name}:
get:
operationId: getChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Get a chute.
description: Retrieve a chute's details by ID or name.
parameters:
- name: chute_id_or_name
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/Chute'
'404':
description: Not found.
put:
operationId: updateChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Update a chute.
description: >-
Update chute metadata such as readme, tagline, logo, and autoscaling
settings.
parameters:
- name: chute_id_or_name
in: path
required: true
schema:
type: string
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/UpdateChuteRequest'
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/Chute'
'422':
description: Validation error.
/chutes/{chute_id}:
delete:
operationId: deleteChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Delete a chute.
description: Remove a chute by ID.
parameters:
- name: chute_id
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
'404':
description: Not found.
/chutes/code/{chute_id}:
get:
operationId: getChuteCode
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Get chute code.
description: Load a chute's source code by ID or name.
parameters:
- name: chute_id
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
content:
text/plain:
schema:
type: string
/chutes/{chute_id}/hf_info:
get:
operationId: getChuteHfInfo
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Get Hugging Face info.
description: >-
Return the Hugging Face repo_id and revision for a chute, used by miners
to pre-download model weights.
parameters:
- name: chute_id
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/HfInfo'
/chutes/warmup/{chute_id_or_name}:
get:
operationId: warmupChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Warm up a chute.
description: Activate a chute (cold start) with an optional quick status check.
parameters:
- name: chute_id_or_name
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
/chutes/utilization:
get:
operationId: getChutesUtilization
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Get utilization.
description: Obtain utilization data from recent capacity logs.
responses:
'200':
description: OK
/chutes/vllm:
post:
operationId: deployVllmChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Deploy a templated vLLM chute.
description: Deploy a templated vLLM model as a chute.
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/DeployTemplateRequest'
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/Chute'
/chutes/diffusion:
post:
operationId: deployDiffusionChute
servers:
- url: https://api.chutes.ai
tags:
- Chutes
summary: Deploy a templated diffusion chute.
description: Deploy a templated diffusion (image generation) model as a chute.
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/DeployTemplateRequest'
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/Chute'
/images/:
get:
operationId: listImages
servers:
- url: https://api.chutes.ai
tags:
- Images
summary: List images.
description: List (and optionally filter/paginate) images.
parameters:
- name: include_public
in: query
required: false
schema:
type: boolean
- name: name
in: query
required: false
schema:
type: string
- name: tag
in: query
required: false
schema:
type: string
- name: page
in: query
required: false
schema:
type: integer
- name: limit
in: query
required: false
schema:
type: integer
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/ImageList'
'422':
description: Validation error.
post:
operationId: createImage
servers:
- url: https://api.chutes.ai
tags:
- Images
summary: Create an image.
description: >-
Store image metadata in the database and kick off an asynchronous image
build that is published to the Chutes registry.
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/CreateImageRequest'
responses:
'202':
description: Accepted - build started.
content:
application/json:
schema:
$ref: '#/components/schemas/Image'
'422':
description: Validation error.
/images/{image_id_or_name}:
get:
operationId: getImage
servers:
- url: https://api.chutes.ai
tags:
- Images
summary: Get an image.
description: Load a single image by ID or name.
parameters:
- name: image_id_or_name
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/Image'
'422':
description: Validation error.
delete:
operationId: deleteImage
servers:
- url: https://api.chutes.ai
tags:
- Images
summary: Delete an image.
description: Delete an image by ID or name:tag.
parameters:
- name: image_id_or_name
in: path
required: true
schema:
type: string
responses:
'200':
description: OK
'422':
description: Validation error.
components:
securitySchemes:
bearerAuth:
type: http
scheme: bearer
description: >-
Chutes API key passed as `Authorization: Bearer cpk_...`. Management
endpoints alternatively accept a Bittensor hotkey signature
(X-Chutes-Hotkey, X-Chutes-Signature, X-Chutes-Nonce).
schemas:
CreateChatCompletionRequest:
type: object
required:
- model
- messages
properties:
model:
type: string
description: Model identifier, e.g. `deepseek-ai/DeepSeek-V3-0324`.
messages:
type: array
items:
$ref: '#/components/schemas/ChatMessage'
max_tokens:
type: integer
temperature:
type: number
format: float
top_p:
type: number
format: float
stream:
type: boolean
default: false
stop:
type: array
items:
type: string
ChatMessage:
type: object
required:
- role
- content
properties:
role:
type: string
enum:
- system
- user
- assistant
- tool
content:
type: string
CreateChatCompletionResponse:
type: object
properties:
id:
type: string
object:
type: string
example: chat.completion
created:
type: integer
model:
type: string
choices:
type: array
items:
type: object
properties:
index:
type: integer
message:
$ref: '#/components/schemas/ChatMessage'
finish_reason:
type: string
usage:
$ref: '#/components/schemas/Usage'
Usage:
type: object
properties:
prompt_tokens:
type: integer
completion_tokens:
type: integer
total_tokens:
type: integer
ListModelsResponse:
type: object
properties:
object:
type: string
example: list
data:
type: array
items:
$ref: '#/components/schemas/Model'
Model:
type: object
properties:
id:
type: string
object:
type: string
example: model
created:
type: integer
owned_by:
type: string
Chute:
type: object
properties:
chute_id:
type: string
name:
type: string
slug:
type: string
tagline:
type: string
readme:
type: string
image:
type: string
public:
type: boolean
created_at:
type: string
format: date-time
ChuteList:
type: object
properties:
items:
type: array
items:
$ref: '#/components/schemas/Chute'
total:
type: integer
page:
type: integer
limit:
type: integer
DeployChuteRequest:
type: object
required:
- name
- image
properties:
name:
type: string
image:
type: string
description: Image ID or name:tag the chute runs from.
code:
type: string
description: Chute source code from the CDK.
node_selector:
type: object
description: GPU / node requirements (e.g. gpu_count, min_vram_gb).
additionalProperties: true
public:
type: boolean
UpdateChuteRequest:
type: object
properties:
readme:
type: string
tagline:
type: string
logo:
type: string
scaling:
type: object
additionalProperties: true
DeployTemplateRequest:
type: object
required:
- model
properties:
model:
type: string
description: Hugging Face model repo to template-deploy.
node_selector:
type: object
additionalProperties: true
public:
type: boolean
HfInfo:
type: object
properties:
repo_id:
type: string
revision:
type: string
Image:
type: object
properties:
image_id:
type: string
name:
type: string
tag:
type: string
status:
type: string
description: Build status (e.g. building, built, error).
public:
type: boolean
created_at:
type: string
format: date-time
ImageList:
type: object
properties:
items:
type: array
items:
$ref: '#/components/schemas/Image'
total:
type: integer
page:
type: integer
limit:
type: integer
CreateImageRequest:
type: object
required:
- name
- tag
properties:
name:
type: string
tag:
type: string
dockerfile:
type: string
description: Dockerfile or build definition for the image.
public:
type: boolean