Vercel Adds Muse Image Generation and Editing to AI Gateway.

Vercel is adding image generation and editing to its AI Gateway through Muse Image, the first image model from Meta Superintelligence Labs. Developers can now call the model using Vercel’s unified API and existing AI SDK workflow, rather than opening a separate provider account or building another integration.

The model accepts both text and image inputs. A developer can send a prompt to create an image from scratch, or provide an existing image with instructions describing how it should be changed. Crucially, both tasks use the same model, reducing the amount of routing logic required inside an application.

Vercel announced the integration on August 26, 2026.

Muse Image is a distinct model family from Muse Spark. While Muse Spark produces text, Muse Image returns images and is designed to handle generation and editing through a single endpoint.

To use it, developers set the model identifier to meta/muse-image-1.0 and call generateImage from the AI SDK. Vercel’s example is compact:

import { generateImage } from 'ai';

const { images } = await generateImage({
  model: 'meta/muse-image-1.0',
  prompt: 'A conference poster for a talk on database indexes.',
});

The response contains the generated images. For an editing workflow, an application can send an image alongside an instruction instead of switching to a separate editing model. That could simplify products such as marketing asset generators, ecommerce image tools, social media editors, and internal design assistants.

For small teams, the operational benefit may matter as much as the model itself. Image AI providers often expose different authentication methods, request formats, response structures, and billing systems. Vercel’s Gateway places Muse Image behind the same access layer used for other supported models.

That provides several practical capabilities:

• One API key, without requiring a separate account with the underlying model provider
• Spend tracking across requests routed through the Gateway
• Request traces for investigating failures and unexpected outputs
• Automatic fallbacks, where supported by the application’s model configuration

The integration therefore gives builders another image option without requiring them to redesign their backend around a provider-specific API.

AI Gateway is becoming a control layer between applications and model providers. Instead of coupling an application directly to each model vendor, developers can use one interface for authentication, observability, and model access. Adding image models broadens that proposition beyond text and transcription workloads.

There are limits to what Vercel has disclosed. The announcement does not provide Muse Image pricing, output dimensions, latency benchmarks, licensing details, or model-specific usage limits. Teams considering the model for production work will need to check the Muse Image model page and current Gateway documentation before committing to a workflow.

The release also fits Vercel’s wider effort to supply infrastructure for AI applications, not just frontend deployment. That push includes controlled model access, execution environments, and agent connectivity. It follows developments such as Vercel Connect reaching general availability with scoped agent access and the Run SDK’s hardened evaluation sandbox.

For builders already using Vercel’s AI SDK, Muse Image should require relatively little integration work. The more consequential change is architectural: image creation and editing can now sit behind the same Gateway controls as the rest of an application’s AI traffic.

What is Vercel Muse Image?

Muse Image is an image generation and editing model from Meta Superintelligence Labs that is available through Vercel AI Gateway. It can create an image from a text prompt or modify an existing image using written instructions, with both tasks handled by the same model.

How do developers access Muse Image through Vercel?

Developers call generateImage from Vercel’s AI SDK and set the model to meta/muse-image-1.0. Requests run through AI Gateway using a Vercel API key, without requiring a separate provider account.

How much does Muse Image cost on Vercel AI Gateway?

Vercel’s announcement does not specify pricing or plan requirements. Developers should consult the current Muse Image model page and AI Gateway documentation for applicable rates and limits.