Google's newest video model, shipped 2026-11-05 and the successor to the Veo 3.1 preview line retired on 2026-10-22. Officially billed per second: $0.07/sec at 720p (30% below 1.1's $0.10); the 1.1 tiers were 360p $0.03 / 1080p $0.15 / 4K $0.30. Google recommends drafting at 360p or 720p with multi-turn edits, then upscaling to 1080p or 4K. Native length rises from 10 to 3–15 seconds, with conversational editing and character / product / ≤3s motion references. A lower per-second rate does not mean a cheaper clip: 15 seconds still costs more than the old 10-second default.
Google, Runway, Luma, xAI and Pika all publish a per-second list price. Google retired the Veo 3.1 preview line on 2026-10-22 in favour of Gemini Omni, so those dead tiers are gone. Figures here come from each vendor’s own pricing page (checked 2026-11); the CNY column uses the PBoC mid rate, not a platform retail mark-up. Subscription-only tools (Midjourney, Adobe Firefly) and the retired Sora 2 API are not listed.
| # | Platform / model | Layer | Price · ¥/sPrice · $/s▲ | vs Seedance 1080P |
|---|---|---|---|---|
| 3 | INTLGemini Omni 1.2 Flash ↗Google · 2026-11-05, 720p $0.07/sec (30% below 1.1) | Overseas | ¥0.47 | ▼ ×0.75 |