whatmodel.sh / video / Veo 3 Fast

Veo 3 Fast

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Veo 3 Fast is a closed-weights video model from Google, released July 2025.

Model information sources

Release date ↗

65.3±1.8comparison only: no current run option verifiedfrom 4 benchmark results across 3 sources

$ how it compares

The video models ranked either side of Veo 3 Fast, out of 139 with a score, then more from Google.

#modelscorewhereRAMreleased
23Wan 2.7 Alibaba68±2api $0.10/sec2026-04
24Vidu Q3 Pro ShengShu (Vidu)65±2api $0.10/sec2026-01
25Veo 3 Fast Google65±2comparison only2025-07
26Runway Aleph 2.0 Runway65±7api $0.28/sec2026-05
27Sora 2 OpenAI65±2comparison only2025-10
more from Google
1Gemini Omni 1.1 Flash Google81±2api $0.10/sec2026-08
3Gemini Omni Flash Google80±2api $0.10/sec2026-05
19Veo 3.1 Google69±1api $0.40/sec2025-10

see all video models · find the ones that fit your machine

$ what it's good at

text to video▇▅▄▆█▆▇▂▃▃▂▄ #14 of 94 · 63% of its score · 3 tests

Beats Grok Imagine Video, Wan 2.7 here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
LMArena video1347.9 ✎▂▂▃▆█▆·▆▅▁▂▇ beats 29/41avg 1147▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 54%typical
VideoVerse overall (Gemini 2.5 Pro judge)1292.03 / 1608 ✎█·▄··▄·····█ beats 4/5easy · avg 1103.17 / 1608▇▇ 5.0%typical
VABench T2AV micro Visual QA81.0% ✎▄····▄····▄█ beats 4/4easy · avg 68%▇▇ 4.7%typical
image to video▁▁▂▂██▄▃▃▄▂▂ #25 of 74 * · 35% of its score · 1 test

Trails Runway Gen-4.5, Kling 2.6 Pro, Wan 2.6 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena image-to-video1326.0 ✎▁▁▁▁▄▇▃▆▃▂▄█ beats 23/40avg 1189▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 35%typical

$ where to run it

No verified API or download recorded yet · add one

$ how the score was worked out · LMArena leaderboard dataset 89%, VideoVerse: Does Your T2V Gen… 5%, VABench: A Comprehensive Benc… 5%

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

sourceresultsshare of the scorelicence
LMArena leaderboard dataset2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 89%CC BY 4.0
VideoVerse: Does Your T2V Generator Have World Model Capability to Synthesize Videos?1▇ 5%CC-BY-4.0
VABench: A Comprehensive Benchmark for Audio-Video Generation (arXiv:2512.09299v2)1▇ 5%CC-BY-4.0
prior · pull toward the middle▇ 1%
weighted average= 65.3

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

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