Imagen 4 API Shutdown

Imagen 4.0 Endpoints Shutting Down

Google has announced the deprecation of all three Imagen 4.0 API model IDs, with a hard shutdown date of August 17, 2026. The affected endpoints are imagen-4.0-generate-001, imagen-4.0-ultra-generate-001, and imagen-4.0-fast-generate-001. After that date, calls to these model IDs will return errors.

All three were released on June 24, 2025 as Google's dedicated image generation family. They offered high-fidelity output, style consistency, and text rendering quality that made them a popular choice for production image generation pipelines.

What Is Being Deprecated

The three affected models cover different performance and quality tiers:

• Imagen 4.0 (imagen-4.0-generate-001): The standard-tier model, optimized for balanced quality and throughput.
• Imagen 4.0 Ultra (imagen-4.0-ultra-generate-001): The high-quality tier, designed for photorealistic output and demanding visual production workflows.
• Imagen 4.0 Fast (imagen-4.0-fast-generate-001): The high-throughput, lower-latency variant for volume-sensitive applications.

Migration to Gemini Native Image Models

This deprecation represents more than a model ID change. Google is consolidating image generation under the Gemini model family, which means the API surface itself changes: instead of calling a dedicated image generation endpoint, developers now use the standard generateContent API with a Gemini image-capable model.

The recommended replacement models are:

• gemini-3.1-flash-image: Google's current recommended replacement for most Imagen 4.0 use cases. Uses the same generateContent endpoint as text-based Gemini calls, accepting text and image inputs and returning generated images in the response.
• gemini-3-pro-image: The higher-quality tier, suitable for production workflows previously using Imagen 4.0 Ultra.

Both models are part of the "Nano Banana" family and support generating images via a unified API call pattern. The shift means image generation becomes a capability of a general-purpose model rather than a dedicated service, which simplifies integration for teams already using the Gemini API for other purposes.

What the Migration Requires

Developers using Imagen 4.0 endpoints will need to:

1. Switch from dedicated Imagen SDK methods or image-generation API calls to the generateContent endpoint.
2. Update model IDs from imagen-4.0-* to gemini-3.1-flash-image or gemini-3-pro-image.
3. Adapt response handling, since generated images are returned within the standard Gemini response structure rather than via a dedicated image response format.

Teams with Canada-hosted deployments should verify regional availability of the replacement models before the August 17 deadline, as Imagen 4.0 has been the only image generation option available in certain geographic configurations.

The official Gemini API deprecations page at ai.google.dev/gemini-api/docs/deprecations provides the authoritative timeline and should be consulted for any updates to the shutdown schedule.