Chutes

Chutes is a permissionless, serverless AI compute platform that lets developers deploy and run any model as an autoscaling "chute" on decentralized GPU capacity (Bittensor Subnet 64). It exposes a single OpenAI-compatible inference endpoint at llm.chutes.ai/v1 for hundreds of open-source LLMs, plus a management REST API at api.chutes.ai for building images and deploying, listing, and operating chutes.

4 APIs 0 Features

AILLMInferenceServerlessGPUBittensor

APIs

Chutes LLM Inference (Chat Completions) API

OpenAI-compatible chat completions across hundreds of open-source models (DeepSeek, Qwen, Llama, GLM, Mistral, and more) served on the Chutes network at a single shared endpoint...

Chutes Image / Other Models API

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 t...

Chutes Management (Deploy / List) API

REST API to list, retrieve, deploy, update, and delete chutes (serverless AI apps), plus templated vLLM/diffusion deploys, warmup, utilization, miner means, and TEE attestation ...

Chutes Images API

Create, list, retrieve, and delete container images that chutes run from; image creation stores metadata and kicks off an asynchronous build that is published to the Chutes regi...

Collections

Chutes API

OPEN

Pricing Plans

Chutes Plans Pricing

5 plans

PLANS

Rate Limits

Chutes Rate Limits

6 limits

RATE LIMITS

FinOps

Chutes Finops

FINOPS

Event Specifications

Chutes Chat Completions Streaming (HTTP + SSE)

AsyncAPI 2.6 description of Chutes' **chat completion streaming** surface. Chutes does not publish a public WebSocket API. The only asynchronous / event-style transport document...

ASYNCAPI

Resources

👥

GitHubOrganization

Sources

opencollection: 1.0.0
info:
  name: Chutes API
  version: '1.0'
request:
  auth:
    type: bearer
    token: '{{bearerToken}}'
items:
- info:
    name: Chat
    type: folder
  items:
  - info:
      name: Create a chat completion.
      type: http
    http:
      method: POST
      url: https://llm.chutes.ai/v1/chat/completions
      body:
        type: json
        data: '{}'
    docs: OpenAI-compatible chat completion. Set stream=true for Server-Sent Events.
- info:
    name: Models
    type: folder
  items:
  - info:
      name: List available models.
      type: http
    http:
      method: GET
      url: https://llm.chutes.ai/v1/models
    docs: Lists the live models currently reachable on the Chutes network.
- info:
    name: Chutes
    type: folder
  items:
  - info:
      name: List chutes.
      type: http
    http:
      method: GET
      url: https://api.chutes.ai/chutes/
    docs: List (and optionally filter/paginate) chutes.
  - info:
      name: Deploy a chute.
      type: http
    http:
      method: POST
      url: https://api.chutes.ai/chutes/
      body:
        type: json
        data: '{}'
    docs: Standard deploy from the Chutes development kit (CDK).
  - info:
      name: Get a chute.
      type: http
    http:
      method: GET
      url: https://api.chutes.ai/chutes/:chute_id_or_name
      params:
      - name: chute_id_or_name
        value: ''
        type: path
        description: The chute ID or name to retrieve.
    docs: Retrieve a chute's details by ID or name.
  - info:
      name: Update a chute.
      type: http
    http:
      method: PUT
      url: https://api.chutes.ai/chutes/:chute_id_or_name
      params:
      - name: chute_id_or_name
        value: ''
        type: path
        description: The chute ID or name to update.
      body:
        type: json
        data: '{}'
    docs: Update chute metadata such as readme, tagline, logo, and scaling settings.
  - info:
      name: Delete a chute.
      type: http
    http:
      method: DELETE
      url: https://api.chutes.ai/chutes/:chute_id
      params:
      - name: chute_id
        value: ''
        type: path
        description: The ID of the chute to delete.
    docs: Remove a chute by ID.
  - info:
      name: Deploy a templated vLLM chute.
      type: http
    http:
      method: POST
      url: https://api.chutes.ai/chutes/vllm
      body:
        type: json
        data: '{}'
    docs: Deploy a templated vLLM model as a chute.
  - info:
      name: Deploy a templated diffusion chute.
      type: http
    http:
      method: POST
      url: https://api.chutes.ai/chutes/diffusion
      body:
        type: json
        data: '{}'
    docs: Deploy a templated diffusion (image generation) model as a chute.
- info:
    name: Images
    type: folder
  items:
  - info:
      name: List images.
      type: http
    http:
      method: GET
      url: https://api.chutes.ai/images/
    docs: List (and optionally filter/paginate) images.
  - info:
      name: Create an image.
      type: http
    http:
      method: POST
      url: https://api.chutes.ai/images/
      body:
        type: json
        data: '{}'
    docs: Store image metadata and kick off an asynchronous image build.
  - info:
      name: Get an image.
      type: http
    http:
      method: GET
      url: https://api.chutes.ai/images/:image_id_or_name
      params:
      - name: image_id_or_name
        value: ''
        type: path
        description: The image ID or name to retrieve.
    docs: Load a single image by ID or name.
  - info:
      name: Delete an image.
      type: http
    http:
      method: DELETE
      url: https://api.chutes.ai/images/:image_id_or_name
      params:
      - name: image_id_or_name
        value: ''
        type: path
        description: The image ID or name:tag to delete.
    docs: Delete an image by ID or name:tag.
bundled: true