whatmodel.sh / image / UniPic 3

UniPic 3

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

UniPic 3 is an open-weights image model from Skywork AI, released January 2026. Licence: mit.

Catalogue identity notes

Skywork UniPic 3 (Wei et al., 2026; Skywork/Unipic3, pinned c1f8544, MIT). Edit-Compass 'Unipic3'. The -DMD and -Consistency-Model distillations are separate repos and are not assumed. Comparison-only.

Model information sources

Release date ↗ · Weights licence ↗

51.8±3.0comparison only: no current run option verifiedfrom 1 benchmark result across 1 source* thinly tested, treat the rank as provisional

Score within the LMArena text-to-image comparison set only.

$ how it compares

The image models ranked either side of UniPic 3, out of 197 with a score, then more from Skywork AI.

#modelscorewhereRAMreleased
93MiniMax image-01 MiniMax53±5api $0.0035/image
94Ideogram 3.0 Ideogram52±2api $0.060/image2025-03
95UniPic 3 Skywork AI52±3comparison only2026-01
96Imagen 3 (imagen-3.0-generate-002) Google52±2comparison only2025-01
97FLUX.2 [klein] 4B Black Forest Labs52±2api $0.0011/image +local~2.9GB2026-01
more from Skywork AI
128UniPic2-SD3.5M-Kontext-2B Skywork AI44±6comparison only2025-08
145UniPic2-Metaquery-9B Skywork AI40±6comparison only2025-08
147Skywork UniPic 1.5B Skywork AI40±6comparison only2025-07

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

$ what it's good at

image to image / editing▁▁▁▃▂▃█▃▃▃▁▁ #59 of 81 * · 96% of its score · 1 test

Beats GPT Image 1 Mini here, though they rank higher overall. Trails GLM-Image, FLUX.2 [klein] 4B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Edit-Compass (EN, overall)47.0% ✎▃▃▂·▃▆▃█·▃·▃ beats 11/20hard · avg 45%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 96%at its level

$ where to run it

No verified API or download recorded yet · add one

Official weights and documentation: huggingface.co/Skywork ↗

$ how the score was worked out · Edit-Compass & EditReward-Com… 96%

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
Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling1▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 96%CC-BY-4.0
prior · pull toward the middle▇ 4%
weighted average= 51.8

$ 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