Skip to main content
This is the API Reference for Prem AI. Here you can find documentation for all of the endpoints that are available in the Prem API. You can use the sidebar to navigate to the different sections. Prem provides a REST API that you can use to interact with the platform with any programming language that can make HTTP requests.

​
🚀 End-to-End Walkthroughs

New to Prem? Start here! These comprehensive walkthroughs guide you through complete workflows from data ingestion to fine-tuning, showing you how to use multiple API endpoints together.

JSONL Dataset Workflow

Upload a JSONL dataset, create snapshots, and launch fine-tuning jobs with your own data.

PDF Synthetic Workflow

Generate synthetic Q&A pairs from PDF documents and fine-tune models on corporate knowledge.

YouTube Synthetic Workflow

Transform YouTube video content into training data and create specialized models.

​
API Endpoints

​
Build your data pipeline

Projects

Provision and inspect project workspaces

Datasets

Ingest JSONL files or trigger synthetic runs

Snapshots

Split datasets into train and validation sets

Recommendations

Request model suggestions for a snapshot

Fine-Tuning

Launch and monitor fine-tuning jobs

​
Run and observe models

Chat Completions

Send prompts to deployed chat models

Models

List all available models

Traces

Retrieve traces captured during inference

​
Authentication

All API endpoints are authenticated using Bearer tokens. You’ll need to create an API key first - see our API Key guide for instructions. Once you have your API key, include it in the Authorization header: 
Authorization: Bearer YOUR_API_KEY
The authentication is automatically picked up from the OpenAPI specification file: 
"security": [
  {
    "bearerAuth": []
  }
]