The three modes
Each mode is tuned for a specific kind of output. You can mix them in one project — generate a hero image, animate a video from that image, then record a voiceover for the reveal.
- 1.
Image
Describe the visual you want. Pick the model that fits the style (fast draft vs. premium photoreal vs. brand illustration). Generate in seconds.
- 2.
Video
Start from a still frame — upload one, or use an image you just generated. Describe the motion, get a short cinematic clip.
- 3.
Speech
Type or paste your script, pick a voice from the library, get studio-grade audio. Good for narration, ads, internal demos.

Choosing the right model
Different models shine at different things. The picker surfaces a small Featured set first (what most teams pick for most jobs), then a full library.
- For marketing photorealism — Nano Banana Pro gives clean studio quality with strong text-in-image.
- For maximum resolution — FLUX.2 Max renders at 4K, best when you need to crop or print.
- For speed + drafts — FLUX.2 Flex generates a usable first pass in under five seconds.
- For video — Kling 2.5 Turbo is fastest today; Seedance 2.0 and Veo 3.1 ship as the premium options when their pipelines stabilize.
Credit cost is variable
Unlike text generation, creative costs depend on the model, resolution, and duration chosen. The picker shows a live credit estimate before you hit Generate so there are no surprise deductions.
| Mode | From | Varies by |
|---|---|---|
| Image | 1 credit | Model tier, resolution |
| Video | 20 credits | Model, duration, resolution |
| Speech | 1 credit | Voice tier, duration |
Variable pricing — the exact credit count is shown in the picker before generation.
Where your assets live
Every generation is saved to your private storage automatically. You can browse the full history from the Creative Studio sidebar, re-download any asset, or drop it directly into a campaign.


