Google closes the Imagen 4 endpoints today
The standard, ultra and fast generate endpoints shut down on 17 August, with callers pointed at the Gemini image model. A generation of image API is being consolidated into the general model.
Google shut down the Imagen 4 standard, ultra and fast generate endpoints on 17 August, directing callers to the Gemini image model instead.
This is consolidation rather than retirement. A dedicated image family is being folded into the general multimodal model, which is the same direction every major lab has moved: one model, many modalities, one endpoint. The separate image API was an artefact of the period when image generation was a different kind of system.
For anyone with production traffic on those endpoints the migration is not free, and the specific cost is prompt behaviour. A general model conditioned to produce images does not respond to the same prompts the same way a dedicated image model did, so a switch that looks like a URL change is a re-evaluation of every prompt in the product.
The structural point is about what an endpoint is worth. Three tiers named standard, ultra and fast were a product taxonomy; a single model with parameters is an argument that the taxonomy was scaffolding. The tiering is moving from the model to the request — in pricing, in reasoning effort, and now in modality.
Digital Applied — AI Model Releases: August 2026 Tracker and Dated Ledger → · LLM Stats — AI Updates Today (August 2026) →