Updated daily at 05:00 UTC
Vision-language
UI-TARS
Open GUI-agent vision-language models that operate computer and phone screens.
By ByteDance.
Versions
4 versions
| Model | Maker | Kind | Released | Context | Input $/1M | Output $/1M |
|---|---|---|---|---|---|---|
| UI-TARS-1.5-7BUI-TARS · Open weights | ByteDance | Vision-language | Apr 2025 | 128K | — | — |
| UI-TARS-2B-SFTUI-TARS · Legacy · Open weights | ByteDance | Vision-language | Jan 2025 | 32.8K | — | — |
| UI-TARS-72B-DPOUI-TARS · Legacy · Open weights | ByteDance | Vision-language | Jan 2025 | 32.8K | — | — |
| UI-TARS-7B-DPOUI-TARS · Legacy · Open weights | ByteDance | Vision-language | Jan 2025 | 32.8K | — | — |
