Benchmark

AI video model 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

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

Clip shown: every model here on the same prompt, “camera-orbit-subject”, 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	Kling 3.0 Turbo Kuaishou		4.84	±0.11	4.72	4.88	4.78	4.84	5.00	—	—	—	40
2	Seedance 2.0 ByteDance		4.84	±0.15	4.79	4.90	4.59	5.00	5.00	—	267	—	39
3	Google Gemini Omni Google		4.82	±0.12	4.81	4.87	4.80	4.81	4.82	—	—	—	40
4	MiniMax H3 MiniMax		4.79	±0.22	4.78	4.86	4.34	5.00	5.00	—	448	—	40
5	Seedance 2.5 ByteDance		4.79	±0.15	4.69	4.88	4.69	4.80	5.00	2.00	233	—	39
6	OpenAI Sora 2 discontinued OpenAI		4.76	±0.17	4.74	4.84	4.25	5.00	5.00	1.00	217	—	40
7	Grok Video xAI		4.72	±0.16	4.70	4.80	4.44	4.71	5.00	1.00	85	—	40
8	Google Veo 3.1 Google		4.63	±0.26	4.59	4.74	4.25	4.81	4.38	—	120	—	40
9	Runway Gen-4.5 Runway		4.57	±0.16	4.48	4.65	4.50	4.65	—	—	—	—	40
10	PixVerse v5.5 PixVerse		4.52	±0.18	4.30	4.68	4.49	4.63	—	—	—	—	40
11	Kling 3.0 Kuaishou		4.47	±0.2	4.31	4.68	3.97	4.82	—	1.00	190	—	40
12	MiniMax Hailuo 02 MiniMax		3.90	±0.47	3.84	4.14	3.59	4.50	0.00	—	300	—	40

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 ranks #1 right now?

As of 23 October 2026, Kling 3.0 Turbo by Kuaishou ranks #1 overall on Genativity's benchmark, scoring 4.84/5 across 40 scored runs (±0.11 at 95% confidence).