Google Imagen4
Google Imagen4 is Kyeo AI's text-only, single-image workflow at 8 credits per request. It accepts a main prompt of up to 5,000 characters, an optional negative prompt, five aspect ratios, and a text seed; the default ratio is 1:1. `Imagen4` is the upstream workflow name, not a quality guarantee.
At a glance
What is this model like?
Kyeo sends the main prompt and any supplied `negative_prompt`, `aspect_ratio`, and text `seed` to `google/imagen4`; it does not send an image. The upstream interface still lists this workflow. Google's direct Imagen 4 Generate workflow is deprecated, and the upstream interface does not say whether it uses that model, another upstream service, or its own parameter conversion.
Key facts
Quickly assess whether this model matches your workflow.
Also known as
The same model may appear under different names across documentation and community discussions; this list keeps them easy to search and compare.
Common search intents
These questions and keywords combine public references with common creative use cases to help you choose faster.
Selection guide
Use these decision points when choosing a model.
Use the workflow when no reference image is needed and one image in any of five fixed ratios is enough for the test.
This entry has no image input, output-count setting, size, format, or safety control. Choose a model that explicitly exposes the required control for source-image preservation, local edits, upscaling, or a specific file format.
Upstream interface caps the negative prompt at 5,000 characters and seed at 500 characters, while Kyeo currently rejects neither overlong value. Hold the main prompt and ratio constant, then change one field at a time.
Popular comparisons
Common comparison questions help you find the practical differences quickly.
Model comparison
Compare the current model with alternatives at a glance.
Practical usage insights
Practical observations based on real creative workflows.
Google Imagen: Fast to explore
Google Imagen: Useful for comparisons
Capabilities
Google Imagen: Clear input options
Google Imagen: Practical controls
Google Imagen: Transparent credit use
Use cases
Ratio comparison from a 1:1 baseline
Negative-prompt and seed comparison
Google Imagen: Option comparison
Prompt tips
Describe visible content in the main prompt
Keep both text parameters within upstream interface limits
Change one variable in seed tests
Strengths
Limitations
Adjustable parameters
Quickly assess whether this model matches your workflow.
Credit usage
Kyeo charges 8 credits for each single-image Google Imagen4 request.
Before batch generation, run an A/B test with the same assets across the current model and alternatives to validate quality and avoid wasting credits.
FAQ
Related models
Compare these similar candidates before deciding.
Google Imagen4 Fast
Google Imagen4 Fast is Kyeo AI's text-only, single-image workflow at 4 credits per request. It accepts a main prompt of up to 5,000 characters, an optional negative prompt, five aspect ratios, and a seed; the default ratio is 16:9. `Fast` is a workflow name, not a measured promise about latency or quality.
Google Imagen4 Ultra
Google Imagen4 Ultra is Kyeo AI's text-only, single-image workflow at 12 credits per request. It accepts a main prompt of up to 5,000 characters, an optional negative prompt, five aspect ratios, and a text seed. `Ultra` is a workflow name, not a Kyeo guarantee that it outperforms the other Imagen4 routes.
4o Image
4o Image is Kyeo AI's text-to-image implementation of the upstream `gpt4o image generation` workflow. It offers `1:1 / 3:2 / 2:3`, costs a fixed 3 credits, and accepts no reference images.
Sources
Content is based on model documentation, feature references, and the settings available on Kyeo AI.
This page treats Kyeo's current configuration, request path, and the upstream interface schema as the available scope. The entry accepts text only, uses `google/imagen4`, limits the main prompt to 5,000 characters, forwards `negative_prompt`, `aspect_ratio`, and a text `seed` when supplied, produces the currently configured single image, and costs 8 credits. The upstream interface also offers an `auto` ratio, caps the negative prompt at 5,000 characters, and caps seed at 500 characters. Kyeo shows five fixed ratios with a 1:1 default but does not yet enforce either text-field limit. Google's direct `imagen-4.0-generate-001` documentation says negative prompts are unsupported, and Google deprecated that workflow on March 24, 2026, with migration recommended before June 30.
Google describes `imagen-4.0-generate-001` as text-to-image with five ratios, PNG or JPEG, up to four direct outputs, and 1K or 2K resolutions, while explicitly listing negative prompts and image editing as unsupported.
Used for Google Imagen lifecycle context; Kyeo's current options remain those shown on the page and in the workbench.
Google's direct guide establishes Imagen 4 family input, output, and lifecycle context.
The upstream schema documents a 5,000-character main prompt and negative prompt, six aspect ratios including `auto`, and a text seed of up to 500 characters; the example workflow is `google/imagen4`.
Kyeo currently accepts text only, limits the main prompt to 5,000 characters, offers five fixed ratios with a 1:1 default, uses a text seed, returns the configured single image, and charges 8 credits.