innFactory AI Consulting from Rosenheim, Germany advises DACH enterprises on GDPR-compliant integration of image generation models. Important notice: Google has shut down the entire Imagen product line. Per Google’s official documentation, Imagen 4 Standard, Fast, and Ultra are no longer usable - the successors are the Gemini image models Gemini 2.5 Flash Image and Gemini 3 Pro Image (Nano Banana Pro). This article documents Imagen for historical reference and lays out Google’s concrete migration guidance. This overview reflects the state as of October 2026.
Imagen has been shut down - what enterprises need to do now
Google has confirmed across several official channels that Imagen as a standalone product has been discontinued:
- Gemini API (ai.google.dev): “Imagen models are shut down. Use Nano Banana for image generation. Imagen is Google’s legacy image generation model. It is now shut down and no longer available in the Gemini API.” Recommended migration:
gemini-2.5-flash-imageorgemini-3.1-flash-imageinstead of the former Imagen model names. - Vertex AI / Gemini Enterprise Agent Platform (cloud.google.com): The Imagen 4 model page lists a “Discontinuation date: June 30, 2026” for
imagen-4.0-generate-001,imagen-4.0-fast-generate-001, andimagen-4.0-ultra-generate-001. A banner at the top of the page reads: “Vertex AI documentation is no longer being updated. Vertex AI’s services are now part of Gemini Enterprise Agent Platform.” - Google’s central Imagen documentation page now describes itself only as a collection of “links to deprecated Imagen pages.”
- Per the Gemini API deprecations table, the concrete shutdown date for
imagen-4.0-generate-001,imagen-4.0-ultra-generate-001, andimagen-4.0-fast-generate-001was August 17, 2026;imagen-3.0-generate-002had already been shut down on November 10, 2025.
One exception is the specialised endpoint imagen-4.0-upscale-preview (pure image upscaling): per Google’s documentation, it is not on the list of discontinued endpoints and was still usable at that time - though it is likewise part of the overall “deprecated” Imagen documentation, so its long-term future is unclear.
Recommended migration
| Former model | Google’s recommendation |
|---|---|
| Imagen 4 Ultra / Standard / Fast | gemini-2.5-flash-image or gemini-3.1-flash-image (Nano Banana / Nano Banana 2) |
| Imagen 3 (all variants) | gemini-2.5-flash-image or gemini-3.1-flash-image |
| imagegeneration@002–@006, imagetext@001 | gemini-2.5-flash-image |
Technically, migration also means an API change per Google: instead of client.models.generate_images (Imagen), use client.models.generate_content (Gemini image models); the response returns image data as content parts instead of a dedicated Imagen response object.
Historical context: what Imagen 4 was
Imagen 4 was introduced in May 2025 as a public preview on Vertex AI and reached GA status in several variants in August 2025:
- Imagen 4 Ultra: highest quality, up to 2K resolution, $0.06 per image
- Imagen 4 (Standard): balanced quality/price, up to 2048×2048 pixels, $0.04 per image
- Imagen 4 Fast: generation times often under 3 seconds, $0.02 per image
All three variants have, per Google, been shut down (see above).
Comparison with the current Gemini image models
| Aspect | Imagen 4 (shut down) | Gemini 2.5 Flash Image | Gemini 3 Pro Image (Nano Banana Pro) |
|---|---|---|---|
| Status | Shut down since mid-2026 | Current (GA) | Current (GA) |
| Focus | Pure text-to-image generation | Image generation + conversational editing | Conversational editing, high quality |
| Resolution | Up to 2K (Ultra) | Standard | Up to 4K |
| Multi-turn editing | No | Yes | Yes |
GDPR availability and migration in the EU
Since Imagen itself is no longer usable, the EU availability question now applies to its successors: Gemini 2.5 Flash Image is, per Google’s current location matrix, available in several EU single regions (including Belgium, Netherlands, Finland, Warsaw, Milan, Madrid - but not Frankfurt); see the Gemini 2.5 Flash Image model page for details. Gemini 3 Pro Image is, per Google, available exclusively via the Global endpoint without guaranteed EU data residency; see the corresponding model page.
Safety and compliance
- SynthID watermark: Was already used by Imagen and is continued by the successor models
- EU AI Act: Generated synthetic media must be labelled as AI-generated - relevant regardless of the model used
- Contract review: Existing integrations that still reference Imagen model names should be checked for functionality immediately
Our recommendation
Existing Imagen integrations should be checked for functionality immediately and migrated to Gemini image models. For most marketing and product workflows, we recommend Gemini 2.5 Flash Image (good price/performance, several EU regions); for maximum quality with 4K output, Gemini 3 Pro Image (Nano Banana Pro) (currently only via the Global endpoint without guaranteed EU data residency).
As alternatives we evaluate Flux (EU-sovereign, Black Forest Labs) and Stability AI depending on use case. Contact us for advice on the right image-model strategy and on migrating existing Imagen workflows.