Overview
imagen-4-fast is Google’s fastest Imagen 4 variant — optimised for quick drafts, social previews, and rapid iteration over hero-fidelity work. Sync API on Vertex AI, blob-hosted output, no async polling.
Lives in the same family as imagen-4 (default) and imagen-4-ultra (premium). Pick by latency + cost; the underlying style language is the same.
Specs
| Field | Value |
|---|---|
| Model ID | imagen-4-fast |
| Creator | |
| Backend | Vertex AI (imagen-4.0-fast-generate-001, us-central1) |
| Best for | Drafts, social previews, rapid iteration |
| Aspect ratios | 1:1 (default), 3:4, 4:3, 16:9, 9:16 |
Sample count n | 1–4 |
| Pricing mode | Flat per image |
| Default latency | ~5–10s |
| Output | Blob-hosted URL (Vercel CDN) |
Pricing
Flat per-image. List = provider cost × 1.35.| Variant | Provider cost | Kyma list |
|---|---|---|
imagen-4-fast | $0.020 | $0.027 |
GET https://kymaapi.com/v1/pricing.
Compared to other image models on Kyma
| Strength | imagen-4-fast | imagen-4 | imagen-4-ultra | nano-banana |
|---|---|---|---|---|
| Cheapest Imagen | ★★★★★ | ★★★★ | ★★ | n/a |
| Photoreal humans | ★★★ | ★★★★ | ★★★★★ | ★★★ |
| Print fidelity | ★★ | ★★★★ | ★★★★★ | ★★ |
| Speed | ★★★★★ | ★★★★ | ★★★ | ★★★★★ |
| Edit-mode | — | — | — | yes |
Use this when
- You need many quick drafts to iterate prompts cheaply.
- The output is a social preview, draft mockup, or low-resolution asset.
- Speed beats fidelity.
Pick something else when
- You need photoreal hero output →
imagen-4orimagen-4-ultra. - You need to edit an existing image (image-in, prompt-driven) →
nano-banana. - You need legible text-in-image →
gpt-image-2(high tier).
Example
202 with a job_id; poll GET /v1/jobs/{id} until status is succeeded. The completed job’s output.url is a Vercel blob URL on Kyma’s CDN.