Netlify Release Notes

Gemini 3.1 Pro Preview Availability through Netlify AI Gateway

Google’s Gemini 3.1 Pro Preview model is now available through Netlify’s AI Gateway with zero configuration required.

Use the Google GenAI SDK directly in your Netlify Functions without managing API keys or authentication. The AI Gateway handles everything automatically. Here’s an example using the Gemini 3.1 Pro Preview model:

import { GoogleGenAI } from '@google/genai';

export default async () => {
  const ai = new GoogleGenAI({});
  const response = await ai.models.generateContent({
    model: 'gemini-3.1-pro-preview',
    contents: 'How can AI improve my workflow?'
  });
  return Response.json(response);
};

Gemini 3.1 Pro Preview is available for all Function types. You get automatic access to Netlify’s caching, rate limiting, and authentication infrastructure.

Learn more in the AI Gateway documentation.

Original source Report a problem

Anthropic’s Claude Sonnet 4.6 on Netlify AI Gateway

Anthropic’s Claude Sonnet 4.6 model is now available through Netlify’s AI Gateway and Agent Runners with zero configuration required.

Use the Anthropic SDK directly in your Netlify Functions without managing API keys or authentication. The AI Gateway handles everything automatically. Here’s an example using the Claude Sonnet 4.6 model:

import Anthropic from '@anthropic-ai/sdk';

export default async () => {
  const anthropic = new Anthropic();
  const response = await anthropic.messages.create({
    model: 'claude-sonnet-4-6',
    max_tokens: 4096,
    messages: [
      {
        role: 'user',
        content: 'How can AI improve my coding?'
      }
    ]
  });
  return Response.json(response);
};

Claude Sonnet 4.6 is available for all Function types and Agent Runners. You get automatic access to Netlify’s caching, rate limiting, and authentication infrastructure.

Learn more in the AI Gateway documentation and Agent Runners documentation.

Original source Report a problem

How it works

You can now sync changes from different agent runs in the Netlify dashboard. This is especially helpful if you use Netlify Drop to publish or update your project and don’t use Git workflows to sychronize versions of your project.

Syncing changes helps you and your whole team build and publish faster.

For example, let’s say you used Netlify Drop to publish your project without setting up a Git workflow.

Next, you decide to use Agent Runners to add a new landing page.

You also start an agent run to update your project’s footer and publish the new footer.

Your agent run for the new landing page doesn’t include the footer changes yet. To get these updates, start a sync run from the Agent Runners dashboard. This will apply all new updates from the live production version of your project.

Now when you publish your landing page updates, you’ll get the updated footer as well.

Syncing changes with Agent Runners enables smoother shipping with several agent runs and with multiple team members working on the same project.

Original source Report a problem

Claude Opus 4.6 available through Netlify’s AI Gateway and Agent Runners

Anthropic’s Claude Opus 4.6 model is now available through Netlify’s AI Gateway and Agent Runners with zero configuration required.
Use the Anthropic SDK directly in your Netlify Functions without managing API keys or authentication. The AI Gateway handles everything automatically. Here’s an example using the Claude Opus 4.6 model:

import Anthropic from '@anthropic-ai/sdk';

export default async () => {
  const anthropic = new Anthropic();
  const response = await anthropic.messages.create({
    model: 'claude-opus-4-6',
    max_tokens: 4096,
    messages: [
      {
        role: 'user',
        content: 'How can AI improve my coding?'
      }
    ]
  });
  return new Response(JSON.stringify(response), {
    headers: {
      'Content-Type': 'application/json'
    }
  });
};

Claude Opus 4.6 is available for all Function types and Agent Runners. You get automatic access to Netlify’s caching, rate limiting, and authentication infrastructure.
Learn more in the AI Gateway documentation and Agent Runners documentation.

Original source Report a problem

Here are some Agent Runners improvements that all people with Credit-based pricing plans can enjoy:

Shareable agent run links

You can now link directly to an agent run to share it with your team or bookmark for later review.

Agent Runners available no matter how you deploy

Agent Runners now supports static projects without build steps. Previously, projects without a build step couldn’t use Agent Runners.

Improved diff view performance

By default, the diff view now loads only the first 50 changed files with an option to load more. This improves performance for large projects.

Feedback welcome

Keep sharing your product feedback about Agent Runners in the feedback form at the bottom of our Agent Runners docs page.

And don’t forget that while you can run multiple agent runs and do other work while they run, you can also play a Netlify game while you wait for the agent to finish.

Original source Report a problem

The Node.js team has released a security update addressing a denial-of-service vulnerability affecting applications that use async_hooks (including in dependencies). Here’s what Netlify customers need to know.

Vulnerability

When async_hooks are enabled on certain versions of Node.js, a stack overflow causes the Node.js process to exit immediately rather than throw a catchable error. This bypasses try-catch blocks and uncaught exception handlers entirely.

A malicious actor could send a crafted payload to crash a server.

Note that many common tools and frameworks use async_hooks under the hood, notably APM and tracing tools (e.g. DataDog, NewRelic, OpenTelemetry) as well as Next.js App Router and other React Server Components implementations. You can find more details on that here.

Impact on Netlify

This is a server-side denial-of-service (DoS) vulnerability. On Netlify, this has minimal impact: our autoscaling serverless architecture means that a malicious request resulting in a crashed or hung function does not affect other requests. However, active exploitation could increase your cold starts and your function costs.

Note that Node.js used during your project’s build is not impacted at all.

What should I do?

If you have opted your Netlify Functions into Node.js 18, we recommend upgrading to Node.js 20 or later. Node.js 18 reached end-of-life in April 2025 and thus will not be patched.

Otherwise, there is no action for you to take. Although this CVE’s impact to Netlify sites is limited, deployed Netlify Functions will be updated to patched Node.js versions automatically on a rolling basis.

For completeness, please note that the Node.js version used during your project’s build is not relevant to this CVE. There is no action for you to take and this will not be automatically patched.

Resources

• Node.js security release announcement
• Netlify Functions Node.js version configuration

Original source Report a problem

Vulnerabilities

The Svelte team has disclosed five CVEs affecting the Svelte and SvelteKit ecosystem. Here’s what Netlify customers need to know.

• CVE-2026-22775: Memory/CPU exhaustion in devalue (5.1.0–5.6.1)
• CVE-2026-22774: Memory exhaustion in devalue (5.3.0–5.6.1)
• CVE-2026-22803: Server crash in @sveltejs/kit (2.49.0–2.49.4)
• CVE-2025-67647: Server crash and SSRF in @sveltejs/kit (2.44.0–2.49.4) and @sveltejs/adapter-node (2.19.0–2.49.4)
• CVE-2025-15265: XSS in svelte (5.46.0–5.46.3)

Impact on Netlify

CVE-2026-22775, CVE-2026-22774, and CVE-2026-22803
These are server-side denial-of-service (DoS) vulnerabilities. On Netlify, these have minimal impact: our autoscaling serverless architecture means that a malicious request resulting in a crashed or hung function does not affect other requests. However, active exploitation could increase your function costs.

In addition, note that only sites using the experimental Remote Functions feature are affected.

CVE-2025-67647
As above, this DoS vulnerability involves intentionally “crashing” a server. On Netlify, this has minimal impact.

The SSRF vulnerability affects a @sveltejs/adapter-node, which is not used by apps deployed to Netlify.

CVE-2025-15265
This is a client-side cross-site scripting (XSS) vulnerability. Regardless of hosting provider, all apps using the experimental hydratable with unsanitized user-controlled keys are vulnerable.

What should I do?

Although the impact to Netlify sites is limited in this case, we always strongly recommend upgrading as soon as possible to patched releases:

• devalue 5.6.2 or later
• @sveltejs/kit 2.49.5 or later
• @sveltejs/adapter-node 5.5.1 or later
• svelte 5.46.4 or later

Resources

Svelte team disclosure

Original source Report a problem

OpenAI GPT-5.2-Codex in Netlify AI Gateway

OpenAI’s GPT-5.2-Codex model is now available through Netlify’s AI Gateway and Agent Runners with zero configuration required.

Use the OpenAI SDK directly in your Netlify Functions without managing API keys or authentication. The AI Gateway handles everything automatically. Here’s an example using the GPT-5.2-Codex model:

import OpenAI from 'openai';

export default async () => {
  const openai = new OpenAI();
  const response = await openai.responses.create({
    model: 'gpt-5.2-codex',
    input: 'How does AI work?'
  });
  return new Response(JSON.stringify(response), {
    headers: {
      'Content-Type': 'application/json'
    }
  });
};

GPT-5.2-Codex is available across Background Functions, Scheduled Functions, and Agent Runners. You get automatic access to Netlify’s caching, rate limiting, and authentication infrastructure.

Learn more in the AI Gateway documentation and Agent Runners documentation.

Original source Report a problem

Prerender.io extension

The third-party Prerender.io service for advanced prerendering is now available as an extension for all Netlify customers.

Previously, you needed a Netlify Pro plan or higher to set up Prerender.io with Netlify’s legacy in-app prerendering feature.

Prerender.io offers advanced configuration options and the Prerender.io dashboard. Learn more about the Prerender.io extension from the extension details page.

If you previously set up Prerender.io using Netlify’s legacy in-app prerendering feature, you need to update your configuration to continue using it this year. Learn more about your next steps and how to check if you need to update your configuration.

To check if your project still needs prerendering, see our prerendering needs checker documentation.

Configuration updates required for existing users

If you have a Netlify Pro plan and you set up the Prerender.io service with Netlify’s legacy prerendering feature (most likely before January 6, 2026), you’ll need to update your configuration to continue using it this year.

To check if your project is using Netlify’s legacy in-app prerendering feature, go to Project configuration > Build & deploy > Post processing > Prerendering from your Netlify project dashboard to see if the legacy prerendering feature is enabled.

Updates are required because Prerender.io no longer relies on Netlify’s legacy in-app prerendering feature, which is being deprecated and will stop working later this year.

Follow our migration steps to update your configuration before these key dates:

Date Impacted plans February 17, 2026 The feature will be disabled for customers on Personal or Pro plans (legacy or credit-based) March 17, 2026 The feature will be disabled for Enterprise and all other specialized plans.

Learn more about this migration process in this migration post.

You can also reach out to Netlify Support for help.

Original source Report a problem