VenusBench-Mobile
Agents use Android apps such as VLC, Markor, and Pro Expense. A model judges 90 primary tasks. The other tasks use programmatic checks.
Ant Group ranks 11 configurations covering ten models on 149 primary tasks across 27 apps. The suite also includes 80 stability variants.
- Made by
- Venus Team, Ant Group
- Released
- February 2026
- Kind
- Phones
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Agent task
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Agentic
- Environment
- Phone apps
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Computer use, Agents
- Strongest model tested
- Kimi K2.5 #76 on the capability index
- Last model added
- January 2026
- Scores refreshed
- November 20th, 2026
Gemini 3 Pro + UI-Venus-72B tops the leaderboard with 36.9% Kimi K2.5 leads with 24.8%
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran on its own with several settings, we keep its best. Agents other companies built around a model get a pale shade of that model's lab color. Gray marks blends and entries not in our model list, such as fine-tunes and older models. None of these count toward leaders. The best model on its own is Kimi K2.5, at 24.8%. Agents built around a model, such as Gemini 3 Pro + UI-Venus-72B, are their own entries. We don't count them as the model's score. This view shows only models we track, each at its best.
- 1 Gemini 3 Pro + UI-Venus-72B Planner-executor 36.9%
- 2 GPT-5.1 + UI-Venus-72B Planner-executor 26.9%
- 3 Kimi K2.5 End-to-end standalone agent 24.8%
- 4 Seed 2.0 End-to-end standalone agent 18.1%
- 5 UI-Venus-72B End-to-end agent 15.4%
- 6 MAI-UI-8B End-to-end agent 12.8%
- 7 MA3 + GUI-Owl-7B Multi-agent framework 12.1%
- 8 Qwen3-VL-30B-A3B End-to-end agent 8.70%
- 9 UI-Venus-7B End-to-end agent 8.10%
- 10 Qwen3-VL-8B End-to-end agent 6.70%
- 11
GUI-Owl-7B End-to-end agent 6.70%
Source: Official VenusBench-Mobile leaderboard
- 1 Kimi K2.5 End-to-end standalone agent 24.8%
- 2
GUI-Owl-7B End-to-end agent 6.70%
Source: Official VenusBench-Mobile leaderboard
| # | Entry | Score | Setting | Date | Link | Source |
|---|---|---|---|---|---|---|
| 1 | Gemini 3 Pro + UI-Venus-72B | 36.9% | Planner-executor | Nov 2025 | Official leaderboard | |
| 2 | GPT-5.1 + UI-Venus-72B | 26.9% | Planner-executor | Nov 2025 | Official leaderboard | |
| 3 | Kimi K2.5 | 24.8% | End-to-end standalone agent | Jan 2026 | Official leaderboard | |
| 4 | Seed 2.0 closed weights | 18.1% | End-to-end standalone agent | Official leaderboard | ||
| 5 | UI-Venus-72B open weights | 15.4% | End-to-end agent | github.com | Official leaderboard | |
| 6 | MAI-UI-8B open weights | 12.8% | End-to-end agent | Official leaderboard | ||
| 7 | MA3 + GUI-Owl-7B | 12.1% | Multi-agent framework | Aug 2025 | Official leaderboard | |
| 8 | Qwen3-VL-30B-A3B open weights | 8.70% | End-to-end agent | Official leaderboard | ||
| 9 | UI-Venus-7B open weights | 8.10% | End-to-end agent | github.com | Official leaderboard | |
| 10 | Qwen3-VL-8B open weights | 6.70% | End-to-end agent | Official leaderboard | ||
| 11 | | 6.70% | End-to-end agent | Aug 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Link | Source |
|---|---|---|---|---|---|---|
| 1 | Kimi K2.5 | 24.8% | End-to-end standalone agent | Jan 2026 | Official leaderboard | |
| 2 | | 6.70% | End-to-end agent | Aug 2025 | Official leaderboard |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official VenusBench-Mobile leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Computer and browser use
- BrowseComp In use Hard-to-find facts that need persistent web browsing. Short verifiable answers graded for exact correctness. Claude Fable 5 93.3%
- ScreenSpot-Pro In use GUI grounding on high-resolution professional-software screenshots. The model must locate the target UI element. GPT-6 Astra 92.7%
- AndroidWorld In use Agents complete 116 tasks across 20 real Android apps in an emulator. Each task checks the device's system state to judge success. GPT-6 Astra 100%
- OSWorld 2.1 In use Agents complete 108 long-horizon workflows across 31 self-hosted websites and desktop apps, averaging 318 tool calls. Scored by binary completion and checkpoint partial credit. Claude Opus 5 44.3%
- WebArena-Verified In use Agents complete 812 audited WebArena tasks on self-hosted websites. Deterministic evaluators check the agent's answer and captured network traces.
Muse Spark 1.1 69%
- OSWorld 2.1 (offline subset) In use Agents complete the 82 OSWorld 2.1 tasks on the official offline list, with no internet access. Scored by checkpoint partial credit and strict pass rate. Claude Fable 5.2 89.3%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 10 months ago, in January 2026, for Kimi K2.5. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the official leaderboard (11 entries).
- Pass Models scored 2 models we track, plus 9 other entries on its leaderboard.
- Warning Newest score The newest came 10 months ago, in January 2026, for Kimi K2.5.
- Note Frontier models None of the top 10 has a score. The strongest tested is Kimi K2.5 (#76).
- Pass Headroom The best is 24.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on ui-venus.github.io: 5 rows for models we track and 6 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.