GPT-OSS LiveGPT-OSS Live
Documentation

Documentation

Everything you need to get started with GPT-OSS

Comprehensive guides, API references, and tutorials to help you deploy and use GPT-OSS models effectively.

Quick Start
Get up and running with GPT-OSS in minutes
1

Install Dependencies

Set up your environment with the required packages

pip install transformers torch accelerate
2

Load the Model

Initialize GPT-OSS model in your application

from transformers import AutoModelForCausalLM, AutoTokenizer
3

Generate Text

Start generating text with your model

output = model.generate(input_ids, max_length=100)

Installation Guide

Multiple ways to install and deploy GPT-OSS models

Ollama
Simplest way to run GPT-OSS locally
  1. Install Ollama from ollama.ai
  2. Run: ollama pull gpt-oss:120b
  3. Start chatting: ollama run gpt-oss:120b
Hugging Face
Use with transformers library
  1. pip install transformers torch
  2. Load model: AutoModelForCausalLM.from_pretrained('openai/gpt-oss-120b')
  3. Generate text with the model
VLLM
High-performance inference server
  1. pip install vllm
  2. vllm serve openai/gpt-oss-120b
  3. Use OpenAI-compatible API
Docker
Containerized deployment
  1. docker pull gpt-oss:120b
  2. docker run -p 8000:8000 gpt-oss:120b
  3. Access via HTTP API
API Reference
Complete API documentation for GPT-OSS models
POST
/v1/chat/completions

Create a chat completion

Parameters:

messages
array
Required
List of messages in the conversation
model
string
Required
Model to use for completion
max_tokens
integer
Maximum number of tokens to generate
temperature
number
Sampling temperature (0-2)
POST
/v1/completions

Create a text completion

Parameters:

prompt
string
Required
Text prompt to complete
model
string
Required
Model to use for completion
max_tokens
integer
Maximum number of tokens to generate

Ready to Build?

Start building amazing applications with GPT-OSS models today.

View ExamplesJoin Community
GPT-OSS Documentation