whatmodel.sh / video / Veo 3
Veo 3
↗ share
+ contribute
Veo 3 is a closed-weights video model from Google, released July 2025. Licence: proprietary.
$ how it compares
The video models ranked either side of Veo 3, out of 139 with a score, then more from Google.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 28 |  Wan 2.6 | 65±1 | api $0.10/sec | ||
| 29 | Wan2.1-VACE-14B | 65±7 | comparison only | ||
| 30 | Veo 3 | 64±2 | comparison only | ||
| 31 |  Kling O3 Pro | 64±4 | api $0.11/sec | ||
| 32 |  Seedance 1.0 Pro Fast | 62±11 | api $0.049/sec | ||
| more from Google | |||||
| 1 |  Gemini Omni 1.1 Flash | 81±2 | api $0.10/sec | ||
| 3 | Gemini Omni Flash | 80±2 | api $0.10/sec | ||
| 19 | Veo 3.1 | 69±1 | api $0.40/sec | ||
$ what it's good at
text to video▇▅▄▆█▆▇▂▃▃▂▄ #16 of 94 * · 51% of its score · 1 test
Beats Wan 2.7, Wan 2.6 here, though they rank higher overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LMArena video | 1340.2 ✎ | ▂▂▃▆█▆·▆▅▁▂▇ beats 26/41 | 51% |
image to video▁▁▂▂██▄▃▃▄▂▂ #29 of 74 · 47% of its score · 3 tests
Trails Wan 2.1 I2V Plus, Runway Gen-4.5, PixVerse V5.6 here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Arena image-to-video | 1331.5 ✎ | ▁▁▁▁▄▇▃▆▃▂▄█ beats 24/40 | 34% |
| PAI-Bench-G human preference arena (Overall ELO) | 1112.2 ✎ | ▄······██··█ beats 6/6 | 11% |
| PAI-Bench-G physical-AI video generation (Overall) | 82.2% ✎ | ▃······▅·█▅█ beats 9/10 | 2.6% |
other tests1 not in a task · 0.9% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| ViGoR-Bench zero-shot visual reasoning, video generators (Result Avg) | 20.5% ✎ | █··█·······█ beats 1/2 | 0.9% |
$ where to run it
No verified API or download recorded yet · add one
$ how the score was worked out · LMArena leaderboard dataset 85%, PAI-Bench: A Comprehensive Be… 13%, ViGoR-Bench: How Far Are Visu… 1%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 LMArena leaderboard dataset | 2 | 85% |
 PAI-Bench: A Comprehensive Benchmark For Physical AI (arXiv:2512.01989v1) | 2 | 13% |
| ViGoR-Bench: How Far Are Visual Generative Models From Zero-Shot Visual Reasoners? (arXiv:2603.25823v1) | 1 | 1% |
| prior · pull toward the middle | 1% | |
| weighted average | = 64.3 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json