‹ PublicAI Index
The LLM benchmark aggregator.
UI-TARS 1.5 7B
ByteDance Seed & Tsinghua University · 7B
Strongest in Computer use (#30 of 37), weakest in General intelligence (#195 of 204). Above par in 0 of 4 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and Kimi K2.5 and ahead of Grok 4.3 and Qwen3.6 27B.
§ 1 · Profile
What it is good at
Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.
Agents44.4−23.7#191/2671/5
Computer use39.9−31.8#30/371/1
Core abilities39.1−28.3#193/2041/3
General intelligence34.3−35.8#195/2041/3
§ 2 · Head to head
What it beats, and what beats it
The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for UI-TARS 1.5 7B, left for the other.
§ 3 · Sources
Where the numbers come from
2 publications, 2 figures. Every one links to the page it was read from.
OSWorld 27.4% (Specialized model)
Kagi LLM Benchmark 20.7%
Badge
[](https://publicai.io/model-index/m/ui-tars-1-5-7b)