01 — Model
02 — Prompt
0/300003 — Settings
Aspect ratio
Resolution
Duration
04 — Start / end frame (optional)
0/2JPG, PNG or WebP up to 10 MB. Only upload images you have the right to use.
Failed runs are refunded automatically. Gemini 4 Carbon is not connected; every run uses the model shown above. Generated files are hosted by our generation provider and kept for at least 24 hours (some video and audio files up to 72 hours). We don't store copies — download anything you want to keep.
05 — Output
Pick a model, describe what you want, and your video will appear here. Each run shows its exact credit cost before you start. Gemini 4 Carbon output never appears here, because Gemini 4 Carbon isn’t connected.
Veo 3.1 Quality at a glance
| Made by | |
|---|---|
| Type | Video generation |
| Settings | 720P / 1080P · 8s · 2 aspect ratios · up to 2 reference images |
| Price on Carbon Studio | from 295 credits per run |
| Outputs per run | 1 |
| Relation to Gemini 4 Carbon | None — a separate model. Gemini 4 Carbon is not connected here. |
Veo 3.1 Quality by Google is available in Carbon Studio today. The highest-fidelity Veo tier for final shots. Pick it in the workbench above, set your options and the exact credit cost appears before you generate. Failed runs are refunded automatically, and results stay online for at least 24 hours.
People often arrive here while reading about Gemini 4 Carbon. To be clear: Veo 3.1 Quality is not Gemini 4 Carbon and is not made from it. Gemini 4 Carbon is Google’s reasoning model in limited preview; see the Gemini 4 Carbon guide. For more video options, visit the video generator.
Other video models
Veo 3.1 Fast
8-second clips with native sound, quick turnaround.
Veo 3.1 Quality
The highest-fidelity Veo tier for final shots.
Veo 3.1 Lite
Low-cost 8-second drafts from text.
Seedance 2.5
Long, coherent motion with audio, up to 15s.
Seedance 2.0
Reliable text- and image-to-video with sound.
Kling 4.0
Cinematic camera work with stereo audio.
Wan 3.0
Budget-friendly clips with audio, 480P to 1080P.
Hailuo 2.3
Expressive characters and camera moves.
MiniMax H3
Multimodal references, 768P or 2K.
Grok Imagine Video 1.5
Playful, fast motion from text or images.
Vidu Q3 Pro
Smooth first-to-last-frame transitions.