Overview

Overview

Build your AI products—right inside GitHub. Create prompts, test models, and ship AI-powered features with built-in tools for model access, prompt collaboration, and lightweight evaluation. Read the docs to learn more.

Get started

Watch the models demo

Watch this 3-minute demo reel to learn everything you can do with GitHub Models.

(Demo video placeholder)

Explore Features

• Create a sample prompt
• Compare multiple prompts
• Compare models

Prompts

Create, evaluate, and iterate on prompts right inside your repo.

0 prompts found
Create a prompt to start building with natural language or using prompt.yml files.

Add AI to your project now

Drop this snippet into your code to start using AI instantly:

import os
from azure.ai.inference import ChatCompletionsClient
from azure.ai.inference.models import SystemMessage, UserMessage
from azure.core.credentials import AzureKeyCredential

endpoint = "https://models.github.ai/inference"
model = "openai/gpt-4o"
token = os.environ["GITHUB_TOKEN"]

client = ChatCompletionsClient(
    endpoint=endpoint,
    credential=AzureKeyCredential(token),
)

response = client.complete(
    messages=[
        SystemMessage("You are a helpful assistant."),
        UserMessage("What is the capital of France?"),
    ],
    temperature=1.0,
    top_p=1.0,
    model=model
)

print(response.choices[0].message.content)

Explore 40+ models in the catalog

Compare models in the playground—test parameters, token usage, and latency to find the right fit for your use case.

Power your prompt with the right model

Test and compare models against your prompt to find the best fit, then commit it directly to your project when you're ready.

Instrument your Actions workflow with models

Set up a new GitHub Actions workflow using models to continually run lightweight evaluations of your AI features on every Pull Request.