Alibaba's Wan 3.0 Tops Artificial Analysis Video Leaderboards with Native Audio
On September 3, Artificial Analysis published leaderboard results for Alibaba's Wan 3.0: it ranked first on the Video Editing (With Audio) leaderboard with 1189 Elo and tied Google Gemini Omni Flash for first on Text to Video at 1237 Elo, while also placing high on Image to Video and other boards. The model supports 1080p, 30-second native audio generation, and official demos confirm its audio-visual sync capability.
Confirmed
- Video Editing (With Audio): Wan 3.0 (Alibaba) ranks first at 1189 Elo (per @ArtificialAnlys, corroborated by @koltregaskes)
- Text to Video: tied for first with Google Gemini Omni Flash at 1237 Elo
- Editing demo: a fish-eye cat POV was restyled into hand-drawn cartoon while preserving first-person motion and sound
- Image-to-video demo: a dragon takes off with synchronized wingbeats, roar, and wind sounds
- Text-to-video demo: a cat taps a toaster's reflection, which taps back, with synced tapping sounds and confused meows
Why it matters
- Audio-visual sync has long been a weakness in video generation. Wan 3.0 demonstrates native audio across editing, text-to-video, and image-to-video tasks while topping multiple leaderboards, making it a leading contender in the current video generation race.
2026-09-03 ~ 2026-09-03 · 6 related posts
Primary sources
- Wan 3.0 ties Gemini Omni Flash for No.1 on Artificial Analysis text-to-video leaderboard — ArtificialAnlys ·
- WAN 3 Tops AI Video Editing Chart, Ranked #1 With Audio at 1189 Elo — koltregaskes ·
- Wan 3.0 demos: fisheye cat POV restyled to hand-drawn cartoon, audio synced — ArtificialAnlys ·
- [source] WAN 3 Tops AI Video Editing Chart, Ranked #1 With Audio at 1189 Elo — koltregaskes · 2026-09-03
- Alibaba's Wan 3.0 debuts #1 in video editing, generates 30s 1080p video with native audio — ArtificialAnlys · 2026-09-03
- Wan 3.0 text-to-video demo: cat meets its distorted toaster reflection, with audio — ArtificialAnlys · 2026-09-03
- Wan 3.0 image-to-video demo: dragon takes flight with synced wingbeats and roars — ArtificialAnlys · 2026-09-03
- [source] Wan 3.0 ties Gemini Omni Flash for No.1 on Artificial Analysis text-to-video leaderboard — ArtificialAnlys · 2026-09-03
- [source] Wan 3.0 demos: fisheye cat POV restyled to hand-drawn cartoon, audio synced — ArtificialAnlys · 2026-09-03