Benchmark

Animation and 3D leaderboard

Board: Overall · Hands · Faces · Liquids · On-screen text · Multi-shot · Lip-sync · Camera · Animals · Style · Image to video · Animation

Look: Realism · Animation and 3D

Animation and 3D adds the anime walk and claymation tests from the core battery to the animation set: 2D cartoon, anime and 3D feature looks. Sora 2 was retired before the animation set was added, so it has no clips on it and is left off this view.

Round: all · v1-vivix-2026-10 · v1-2026-10 · v1-2026-09 · anim1-2026-10

Clip shown: every model here on the same prompt, “anim-walk-cycle”, so the preview is a fair comparison, not a cherry-picked best clip.

#	Model	Clip	Overall	95% CI	Fidelity	Visual	Physics	Consistency	Audio	Attempts	Gen (s)	$/run	Runs
1	Google Veo 3.1 Google		4.85	±0.19	4.88	4.88	4.63	5.00	—	—	—	—	4
2	Google Gemini Omni Google		4.85	±0.25	4.75	5.00	4.88	4.75	—	—	—	—	4
3	Seedance 2.0 ByteDance		4.72	±0.44	4.13	5.00	4.88	4.88	—	2.00	—	—	4
4	MiniMax H3 MiniMax		4.69	±0.74	4.63	4.88	4.50	4.75	—	—	—	—	4
5	FLUX 3 Black Forest Labs		4.63	±0.49	4.50	4.88	4.38	4.75	—	—	—	—	4
6	Vivix W1 preview Vivix		4.57	±0.5	4.13	4.88	4.63	4.63	—	1.00	20	$0.026	4
7	Grok Video xAI		4.57	±0.35	4.50	5.00	4.25	4.50	—	—	—	—	4
8	Wan 3.0 preview Alibaba		4.53	±0.62	4.25	5.00	4.25	4.63	—	—	—	—	4
9	Kling 4.0 preview Kuaishou		4.53	±0.52	4.50	4.88	4.13	4.63	—	—	—	—	4
10	Kling 3.0 Kuaishou		4.50	±0.63	4.13	4.88	4.38	4.63	—	—	—	—	4
11	Seedance 2.5 ByteDance		4.44	±0.74	4.00	4.88	4.00	4.88	—	2.00	—	—	4
12	Kling 3.0 Turbo Kuaishou		4.41	±0.57	4.25	4.63	4.25	4.50	—	—	—	—	4
13	Runway Gen-4.5 Runway		4.25	±0.86	4.38	4.50	4.25	3.88	—	—	—	—	4
14	MiniMax Hailuo 02 MiniMax		4.13	±0.58	3.88	4.50	4.00	4.13	—	—	—	—	4
15	PixVerse v5.5 PixVerse		3.82	±0.47	3.25	4.38	3.50	4.13	—	—	—	—	4

Attempts is the average number of generations needed before one was usable, the number that actually decides what a model costs you. 95% CI is computed from that model's own runs on this board, not assumed. On a category board with only 4 runs per model, a wide CI is the honest reflection of a small sample, not a bug. Two models whose CIs overlap are not meaningfully different at this sample size.

Frequently asked

Which AI video model is best at Animation and 3D right now?

As of 23 October 2026, Google Veo 3.1 by Google ranks #1 on Animation and 3D on Genativity's benchmark, scoring 4.85/5 across 4 scored runs (±0.19 at 95% confidence).