Imagen 4 Ultra by Google DeepMind — highest quality text-to-image model. Fine-grained textures, accurate colours, photorealistic compositions. C2PA watermarking. $0.06/image.
Three variants for different quality and speed trade-offs — all on the same Runflow API.
Highest quality — maximum detail, sharpness, and realism
Balanced quality and speed for production workloads
Sub-second generation optimised for real-time apps
Every model on Runflow is scored per vertical — not generic leaderboard numbers. Scores come from domain-expert evaluation on real production prompts.
| Use Case | Assessment | Notes |
|---|---|---|
| Portrait | Strong | Highest quality photorealism with fine-grained textures |
| Fashion | Strong | Accurate colours and editorial-level detail |
| Product | Strong | Photorealistic compositions with C2PA watermarking |
| Creative | Strong | Top-tier artistic flexibility and composition |
Full parameter reference for the imagen4-preview-ultra model endpoint.
| Parameter | Type | Description |
|---|---|---|
| prompt | string | Text description of the image to generate. |
| resolution | string | Output resolution. Options: "1K", "2K". Default: "1K". |
| aspect_ratio | string | Image aspect ratio. Options: "1:1", "16:9", "9:16", "4:3", "3:4". Default: "1:1". |
| num_images | integer | Number of images to generate. Range: 1–4. Default: 1. |
| output_format | string | Output format. Options: "jpeg", "png", "webp". Default: "png". |
| safety_tolerance | string | Content moderation strictness. 1 = most strict, 6 = least strict. Default: 4. |
| seed | integer | Random seed for reproducible outputs. |
| sync_mode | boolean | Return media as data URI without storing in request history. Default: false. |
Flux AI models are available on many platforms. Here's what makes the Runflow image generation API different.
Imagen 4 Ultra is Google DeepMind's flagship generation model — optimised for maximum sharpness, detail, and photorealism.
Every image is digitally watermarked with C2PA metadata, enabling provenance verification — important for editorial, legal, and brand-safety use cases.
Generate from prompts in English, Chinese, Hindi, Japanese, Korean, Portuguese, or Spanish — no translation step needed.
Runflow's Sentinel system scores every output before delivery, catching low-quality or off-prompt results automatically.
Same API, different strengths. Switch models with one parameter change.
Black Forest Labs
FLUX.1 [dev] is a 12B-parameter text-to-image model by Black Forest Labs. Open-w…
ByteDance
Seedream 4.5 by ByteDance — unified generation and editing in one model. Strong…
ByteDance
Seedream 4 by ByteDance — unified generation and editing model with photorealism…
Nano Banana 2 by Google — fast image generation across 4 resolution tiers (0.5K–…
OpenAI
GPT-Image 1.5 by OpenAI — high-fidelity images with strong prompt adherence. Thr…
Nano Banana Pro by Google — premium image generation with sharp detail and natur…
Get API access in minutes. No GPU setup, no infrastructure to manage. Pair with Sentinel to control the quality of every AI-generated image you ship.