Back to API Docs
📡 COMPUTE API

Griddly Compute API

Access distributed GPU compute programmatically. Run AI inference, training, and batch processing at scale.

Quick Start

Base URL:

https://api.griddly.ai/v1

Authentication:

API Key in header

Example Request:

curl -X POST https://api.griddly.ai/v1/tasks \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"type": "inference", "model": "llama-70b", "input": "Hello world"}'

Endpoints

POST /v1/tasks

Create new compute task

POST /v1/tasks

GET /v1/tasks/{id}

Get task status

GET /v1/tasks/{id}

GET /v1/tasks

List all tasks

GET /v1/tasks

DELETE /v1/tasks/{id}

Cancel task

DELETE /v1/tasks/{id}

GET /v1/models

List available models

GET /v1/models

GET /v1/usage

Get usage statistics

GET /v1/usage

GET /v1/billing

Get billing info

GET /v1/billing

Task Types

inference

Run model inference

training

Fine-tune models

batch

Batch processing

custom

Custom ONNX models

Response Codes

CodeDescription
200Success
201Created
400Bad Request
401Unauthorized
429Rate Limited
500Server Error

Rate Limits

Starter

100 req/min

Growth

500 req/min

Scale

2000 req/min

Enterprise

Unlimited

SDKs

Python

pip install griddly

Node.js

npm install @griddly/sdk

Coming Soon

Go, Rust

Ready to Get Started?

Get your API key and start building with Griddly Compute API.