Model reference · open weights
UI-TARS-DPO is an open-weight language model from ByteDance. UI-TARS-7B-DPO (BF16) weighs 16.6 GB; the smallest configuration that runs it is 2× RTX 3060 12 GB.
Summary of the ByteDance-Seed/UI-TARS-7B-DPO model card, 2026-10-01
What it is
| Released by | ByteDance |
|---|---|
| Released | 2025-01-22 |
| Parameters | 8.3B |
| VRAM | 16.6 GB for the weights |
What it runs on
| Card | Requests at once | Context max | Memory | |
|---|---|---|---|---|
| 8K each | 32K each | |||
| RTX 3060 12 GB … RTX 4060 Ti 16 GB 2 smaller cards | — | — | — | |
| RTX 3090 24 GB | 10 | 2 | all 32K | 23.4 GB |
| RTX 4090 24 GB | 9 | 2 | all 32K | 23.4 GB |
| RTX 5090 32 GB | 26 | 6 | all 32K | 31.0 GB |
| L40S 48 GB | 53 | 13 | all 32K | 44.0 GB |
| A100 80 GB | 126 | 31 | all 32K | 78.2 GB |
| H100 80 GB | 117 | 29 | all 32K | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | 150 | 37 | all 32K | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | 179 | 44 | all 32K | 107 GB |
| H200 141 GB | 244 | 61 | all 32K | 138 GB |
| B200 180 GB | 324 | 81 | all 32K | 176 GB |
| 2× RTX 3060 12 GB tensor parallel | 5 | 1 | all 32K | 11.6 GB a card |
| 2× RTX 4060 Ti 16 GB tensor parallel | 21 | 5 | all 32K | 15.4 GB a card |
| Requests at once | 8K tokens each | 32K tokens each |
|---|---|---|
| 1 | 19.1 GB | 20.5 GB |
| 5 | 21.0 GB | 28.1 GB |
| 8 | 22.4 GB | 33.7 GB |
| 16 | 26.2 GB | 48.7 GB |
| 32 | 33.7 GB | 78.8 GB |
| 64 | 48.7 GB | 139 GB |
One card, with vLLM's small-card settings.
From the model card
UI-TARS-2B-SFT | UI-TARS-7B-SFT | UI-TARS-7B-DPO(Recommended) | UI-TARS-72B-SFT | UI-TARS-72B-DPO(Recommended)
UI-TARS is a next-generation native GUI agent model designed to interact seamlessly with graphical user interfaces (GUIs) using human-like perception, reasoning, and action capabilities. Unlike traditional modular frameworks, UI-TARS integrates all key components—perception, reasoning, grounding, and memory—within a single vision-language model (VLM), enabling end-to-end task automation without predefined workflows or manual rules.
This repository contains the model for the paper UI-TARS: Pioneering Automated GUI Interaction with Native Agents.
Code: https://github.com/bytedance/UI-TARS
Perception Capabilty Evaluation
| Model | VisualWebBench | WebSRC | SQAshort |
|---|---|---|---|
| Qwen2-VL-7B | 73.3 | 81.8 | 84.9 |
| Qwen-VL-Max | 74.1 | 91.1 | 78.6 |
| Gemini-1.5-Pro | 75.4 | 88.9 | 82.2 |
| UIX-Qwen2-7B | 75.9 | 82.9 | 78.8 |
| Claude-3.5-Sonnet | 78.2 | 90.4 | 83.1 |
| GPT-4o | 78.5 | 87.7 | 82.3 |
| UI-TARS-2B | 72.9 | 89.2 | 86.4 |
| UI-TARS-7B | 79.7 | 93.6 | 87.7 |
| UI-TARS-72B | 82.8 | 89.3 | 88.6 |
Grounding Capability Evaluation
| Agent Model | Dev-Text | Dev-Icon | Dev-Avg | Creative-Text | Creative-Icon | Creative-Avg | CAD-Text | CAD-Icon | CAD-Avg | Scientific-Text | Scientific-Icon | Scientific-Avg | Office-Text | Office-Icon | Office-Avg | OS-Text | OS-Icon | OS-Avg | Avg-Text | Avg-Icon | Avg |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| QwenVL-7B | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.7 | 0.0 | 0.4 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.1 |
| GPT-4o | 1.3 | 0.0 | 0.7 | 1.0 | 0.0 | 0.6 | 2.0 | 0.0 | 1.5 | 2.1 | 0.0 | 1.2 | 1.1 | 0.0 | 0.9 | 0.0 | 0.0 | 0.0 | 1.3 | 0.0 | 0.8 |
| SeeClick | 0.6 | 0.0 | 0.3 | 1.0 | 0.0 | 0.6 | 2.5 | 0.0 | 1.9 | 3.5 | 0.0 | 2.0 | 1.1 | 0.0 | 0.9 | 2.8 | 0.0 | 1.5 | 1.8 | 0.0 | 1.1 |
| Qwen2-VL-7B | 2.6 | 0.0 | 1.3 | 1.5 | 0.0 | 0.9 | 0.5 | 0.0 | 0.4 | 6.3 | 0.0 | 3.5 | 3.4 | 1.9 | 3.0 | 0.9 | 0.0 | 0.5 | 2.5 | 0.2 | 1.6 |
| OS-Atlas-4B | 7.1 | 0.0 | 3.7 | 3.0 | 1.4 | 2.3 | 2.0 | 0.0 | 1.5 | 9.0 | 5.5 | 7.5 | 5.1 | 3.8 | 4.8 | 5.6 | 0.0 | 3.1 | 5.0 | 1.7 | 3.7 |
| ShowUI-2B | 16.9 | 1.4 | 9.4 | 9.1 | 0.0 | 5.3 | 2.5 | 0.0 | 1.9 | 13.2 | 7.3 | 10.6 | 15.3 | 7.5 | 13.5 | 10.3 | 2.2 | 6.6 | 10.8 | 2.6 | 7.7 |
| CogAgent-18B | 14.9 | 0.7 | 8.0 | 9.6 | 0.0 | 5.6 | 7.1 | 3.1 | 6.1 | 22.2 | 1.8 | 13.4 | 13.0 | 0.0 | 10.0 | 5.6 | 0.0 | 3.1 | 12.0 | 0.8 | 7.7 |
| Aria-UI | 16.2 | 0.0 | 8.4 | 23.7 | 2.1 | 14.7 | 7.6 | 1.6 | 6.1 | 27.1 | 6.4 | 18.1 | 20.3 | 1.9 | 16.1 | 4.7 | 0.0 | 2.6 | 17.1 | 2.0 | 11.3 |
| UGround-7B | 26.6 | 2.1 | 14.7 | 27.3 | 2.8 | 17.0 | 14.2 | 1.6 | 11.1 | 31.9 | 2.7 | 19.3 | 31.6 | 11.3 | 27.0 | 17.8 | 0.0 | 9.7 | 25.0 | 2.8 | 16.5 |
| Claude Computer Use | 22.0 | 3.9 | 12.6 | 25.9 | 3.4 | 16.8 | 14.5 | 3.7 | 11.9 | 33.9 | 15.8 | 25.8 | 30.1 | 16.3 | 26.9 | 11.0 | 4.5 | 8.1 | 23.4 | 7.1 | 17.1 |
| OS-Atlas-7B | 33.1 | 1.4 | 17.7 | 28.8 | 2.8 | 17.9 | 12.2 | 4.7 | 10.3 | 37.5 | 7.3 | 24.4 | 33.9 | 5.7 | 27.4 | 27.1 | 4.5 | 16.8 | 28.1 | 4.0 | 18.9 |
| UGround-V1-7B | - | - | 35.5 | - | - | 27.8 | - | - | 13.5 | - | - | 38.8 | - | - | 48.8 | - | - | 26.1 | - | - | 31.1 |
| UI-TARS-2B | 47.4 | 4.1 | 26.4 | 42.9 | 6.3 | 27.6 | 17.8 | 4.7 | 14.6 | 56.9 | 17.3 | 39.8 | 50.3 | 17.0 | 42.6 | 21.5 | 5.6 | 14.3 | 39.6 | 8.4 | 27.7 |
| UI-TARS-7B | 58.4 | 12.4 | 36.1 | 50 |
Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.