whatmodel.sh / video / Ray 3

Ray 3

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

Ray 3 is a closed-weights video model from Luma AI, released September 2025. Licence: proprietary.

54.5±1.8comparison only: no current run option verifiedfrom 3 benchmark results across 2 sources

$ how it compares

The video models ranked either side of Ray 3, out of 139 with a score, then more from Luma AI.

#modelscorewhereRAMreleased
47Vidu Q2 Turbo ShengShu (Vidu)55±3api $0.050/sec2025-09
48Hailuo 2.3 MiniMax55±2api $0.047/sec2025-10
49Ray 3 Luma AI55±2comparison only2025-09
50Vidu Q3 Shengshu54±12comparison only2026-01
51Kling 2.1 Standard Kuaishou (Kling AI)54±3api $0.050/sec2025-05
more from Luma AI
66Luma Dream Machine 1.6 Luma AI50±7comparison only2024-09
94Ray 2 Luma AI43±2comparison only2025-01
97Luma Dream Machine 1.0 Luma AI42±7comparison only2024-06

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

$ what it's good at

text to video▇▅▄▆█▆▇▂▃▃▂▄ #28 of 94 * · 60% of its score · 2 tests

Beats Seedance 1.0 Pro, Kling O1 Pro, P-Video here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
LMArena video1206.2 ✎▂▂▃▆█▆·▆▅▁▂▇ beats 16/41avg 1147▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 58%typical
HumanScore (T2V)87.7% ✎▂··▂··█▄▃··▆ beats 8/11easy · avg 84%▇ 1.8%blunt
image to video▁▁▂▂██▄▃▃▄▂▂ #45 of 74 * · 38% of its score · 1 test

Trails Cosmos3-Nano, Kling 2.1 Master, Hailuo 02 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena image-to-video1225.9 ✎▁▁▁▁▄▇▃▆▃▂▄█ beats 10/40avg 1189▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 38%typical

$ where to run it

No verified API or download recorded yet · add one

$ how the score was worked out · LMArena leaderboard dataset 97%, HumanScore: Benchmarking Huma… 2%

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▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 97%CC BY 4.0
HumanScore: Benchmarking Human Motions in Generated Videos1▇ 2%CC-BY-4.0
prior · pull toward the middle▇ 1%
weighted average= 54.5

$ 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