Model Overview
A preview version of the standard text-to-image model in Google's Imagen 4 series. It takes text prompts as input and outputs images; it is distinct from the Fast and Ultra versions. (developers.googleblog.com)
Capabilities and Uses
- Generates images in different styles from text descriptions, including photorealistic and abstract styles. Google says Imagen 4 improves the spelling and typography of text in images compared with previous models. (developers.googleblog.com)
- Can be used to create visual content such as posters, greeting cards, comics, and presentation materials. The Imagen 4 series supports multiple aspect ratios and resolutions up to 2K; the complete specifications of this particular preview endpoint have not been verified. (blog.google)
- Google says images generated by Imagen 4 contain an invisible SynthID digital watermark. (developers.googleblog.com)
Limitations and Status
- This is a preview endpoint for image generation; image editing capabilities or parameters of other versions should not be assumed to apply to this version. The full functional limitations of the preview endpoint have not been verified.
- In an October 2025 announcement, Vertex AI said the preview endpoint was scheduled for removal on November 30, 2025; a later announcement still listed it as a deprecated endpoint and recommended migration before March 19, 2026. The two announcements have different timelines and should not be used to infer the exact date of actual shutdown. (docs.cloud.google.com)
- The Gemini API deprecation table lists this model as deprecated, with a shutdown date of February 17, 2026, and currently recommends
gemini-nano-banana-2.1as the replacement. Lifecycle announcements for different platforms should be considered separately. (aistudio.google.com)
Release Date
The Vertex AI release notes announced on June 11, 2025 that the Imagen 4 public preview model was updated to this ID. The Gemini API deprecation table lists June 24, 2025 as the release date for the same ID; the Google Developers Blog also announced the Gemini API paid preview on that date. (docs.cloud.google.com)
Pricing
- Historical Gemini API preview price: On June 24, 2025, Google announced a price of 0.04 US dollars (USD) per output image for standard Imagen 4; the input was a text prompt, and the announcement did not list a separate input charge. This was the preview price at the time and is not a commitment regarding the current availability or current price of the deprecated endpoint. (developers.googleblog.com)
- Vertex AI: Historical input and output charges for this specific preview ID have not been verified. The current Google Cloud pricing page lists a generation price of 0.04 USD per image for text input and image output under the Imagen 4 series, but does not explicitly state that the entry applies to this deprecated preview ID. It is therefore not treated as the historical price for this endpoint. (cloud.google.com)
- Date pricing sources were consulted: November 18, 2026.