Netlify AI Gateway

What it does

AI Gateway takes requests made with the official OpenAI, Anthropic, Google Gemini, or OpenRouter SDKs from code running on Netlify, such as serverless, edge, or background functions, and routes them to the chosen model. Netlify injects the required API keys and base URLs at runtime, so developers do not configure keys or separate provider accounts. The output is the model's normal response, such as chat completions, messages, or generated content, returned through the same SDK calls the developer already uses.

People use it to add AI features to a site or app they deploy on Netlify, switching models without rewriting code. The distinct parts are the built-in key handling, unified usage analytics in the Netlify dashboard, per-account rate limits, and a claim that prompts and model outputs are not stored. Local development is supported through the Netlify CLI or Vite plugin.

It suits developers already building on Netlify. Access requires a Netlify account; a Free plan is listed, with usage billed in credits and AI inference costs varying by model. Key limitations: AI inference cost depends on the model used, free-tier usage is capped at 300 credits per month with projects pausing at limits, and paid use requires credits or auto recharge.

What it can do

• Call OpenAI models through the official OpenAI SDK without managing an API key or base URL.
• Call Anthropic models through the official Anthropic SDK with keys injected by Netlify.
• Call Google Gemini models through the official Google GenAI SDK with keys injected by Netlify.
• Call open models such as Kimi and DeepSeek, plus many more, through the OpenRouter SDK.
• View token usage and credit consumption per account in the Netlify dashboard.
• Apply per-account rate limits to production traffic across projects.

Works with

• OpenAI SDK
• Anthropic SDK
• Google GenAI SDK
• OpenRouter SDK
• Netlify CLI
• Vite plugin

Before you choose

• AI inference costs vary by model and are metered in credits; check the pricing page rate table before estimating spend.
• On the Free plan, usage is capped at 300 credits per month and projects pause at limits; auto recharge is off by default on paid plans.
• Netlify states prompts and model outputs are not stored, but governance and data handling for your own requirements should be reviewed.

Listed under

• Coding
• Sandboxes and infrastructure

Get started

Start with the free plan and enable AI Gateway in a Netlify project.

1. Create a Netlify account at app.netlify.com/signup or start from an existing Netlify project.
2. Install the official SDK for your chosen provider (OpenAI, Anthropic, Google GenAI, or OpenRouter) in your project.
3. Call the model from a Netlify function using the SDK; Netlify injects the API key and base URL at runtime, and the local CLI can be used for development.

Free plan includes 300 credits per month with hard limits. Paid plans bill credits; AI inference costs are calculated per model and are not listed as fixed rates on the page. Auto recharge is off by default and only Team Owners can enable it. Enterprise pricing is custom.

Source: netlify.com/pricing, checked 30 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 30 Oct 2026. Each source supports part of the overview; availability and prices can change.

• AI Gateway | Netlify ↗www.netlify.com
• Pricing and Plans | Netlify ↗www.netlify.com

Company website

netlify.com: product, pricing and policy pages

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 30 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.