Imagen 4 (Fast / Standard / Ultra).
Google's photorealistic text-to-image model.

Coinis offers three Imagen 4 tiers. Fast is $0.034 per image. Standard is $0.068 per image. Ultra is $0.102 per image. All prices are pay-as-you-go with no minimum commitment.

Fast is optimized for speed and volume. It is the lowest-cost tier at $0.034/image and does not support the 2K resolution option. Standard adds 1K-to-2K resolution control at $0.068/image and suits most production workflows. Ultra is the highest-fidelity tier at $0.102/image, designed for hero shots and campaigns where output quality is the top priority.

On Coinis, Imagen 4 (Fast / Standard / Ultra) is pay-as-you-go from one shared token wallet. Buy tokens once. Spend them on any model. No separate accounts. No monthly commit.

Imagen 4 is Google DeepMind's photorealism-first model. It includes built-in SynthID provenance watermarking and excels at true-to-life detail, fine textures, and readable in-image text. Flux 1.1 Pro is a strong alternative for teams that prioritize stylistic range or open-weights flexibility. If enterprise provenance and photorealism are the priorities, Imagen 4 is the closer fit.

Not through the Gemini Developer API path. SynthID watermarking is on by default and cannot be disabled via the Gemini API. To get watermark-free output, use the Vertex AI API and set the addWatermark parameter to false. Note that seed-based deterministic generation also requires watermarking to be disabled.

Yes. Imagen 4 Standard and Ultra both support 2K output (up to 2048px) via the imageSize parameter. Imagen 4 Fast does not support the imageSize configuration and outputs at a fixed resolution. Choose Standard or Ultra when 2K output is required.

Yes, text-in-image rendering is a supported capability. For best results, keep each text phrase to 25 characters or fewer and use no more than 2 to 3 distinct phrases per image. Text rendering is still an evolving feature and may require a few regeneration attempts for complex layouts.