Netlify Adds OpenAI GPT-5.6 Models to AI Gateway With Zero Configuration.

Netlify announced that OpenAI's GPT-5.6 models—Sol, Luna, and Terra—are now available through its AI Gateway and Agent Runners. Developers can call these models directly from Netlify Functions using the OpenAI SDK without managing API keys or wiring up authentication.

The update eliminates a common friction point in serverless AI development. Typically, developers need to store API keys as environment variables, handle authentication manually, and implement rate limiting and caching separately. Netlify's AI Gateway handles all of this automatically.

Here's the complete code needed to use GPT-5.6 Sol in a Netlify Function:

import OpenAI from 'openai';

export default async () => {
    const openai = new OpenAI();
    const response = await openai.responses.create({
        model: 'gpt-5.6-sol',
        input: 'Give a concise explanation of how AI works.',
    });
    return Response.json(response);
};

No configuration files, no environment variables, no authentication setup. The AI Gateway detects the OpenAI SDK calls and routes them through Netlify's infrastructure.

The AI Gateway provides caching, rate limiting, and authentication infrastructure automatically. When a Function calls an OpenAI model, Netlify handles token management and response caching based on request patterns. This reduces latency for repeated queries and helps manage costs by avoiding duplicate API calls.

Rate limiting prevents Functions from exceeding OpenAI's API quotas. The Gateway tracks usage across all Functions in a project and applies throttling when necessary. Developers don't need to implement circuit breakers or retry logic—the infrastructure layer handles it.

All three GPT-5.6 models work across every Function type Netlify supports, including background functions, scheduled functions, and edge functions. They also work in Agent Runners, Netlify's environment for running AI coding agents, following the company's recent redesign of Functions for agent-first development.

OpenAI's GPT-5.6 line includes three models optimized for different use cases:

Sol prioritizes speed with lower latency and reduced token costs. It handles straightforward queries where response time matters more than nuanced reasoning.

Luna balances performance and capability for general-purpose applications. It provides stronger reasoning than Sol while maintaining reasonable response times.

Terra offers the most advanced reasoning and context handling in the series. It's designed for complex queries that require deep analysis or long-context understanding.

Developers can switch between models by changing the model parameter in their OpenAI SDK calls. The AI Gateway routes requests to the appropriate model without additional configuration.

The GPT-5.6 models are available immediately for all Netlify users. No account changes or feature flags are required. Existing Functions can start using the models by updating their OpenAI SDK calls and deploying.

Netlify's AI Gateway documentation covers authentication flows, caching behavior, and rate limit configuration. The Agent Runners documentation explains how to use the models in AI coding workflows.

The update continues Netlify's push into AI-native infrastructure. The company has been adding agent-focused features to its platform, including per-function compute controls and cross-platform function interfaces. The AI Gateway integration removes another layer of configuration complexity, letting developers focus on building AI features rather than managing infrastructure.