Benchmark Almanac

GUI Knowledge Bench

o3 78.9% leads since October 2025
Not in model cards

Models answer 3,483 questions about interface states, actions, and procedures. They inspect screenshots and task sequences from 292 applications across six platforms.

Alibaba reports 75.5% for GUI-Owl 1.5. The makers’ online leaderboard ranks nine models by the unweighted mean of eight scores.

Released
October 2025
Kind
Finding things on screen
Level
Everyday
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Screens → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
Strongest model tested
o3 #84 on the capability index
Last model added
August 2025
Scores refreshed
November 20th, 2026
9
Entries on its own leaderboard, 8 models we track

o3 leads with 78.9% among models released by August 2025 o3 leads with 78.9% among models released by August 2025

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in August 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 o3 78.9%
  2. 2 Gemini 2.5 Pro Jun 2025 77.1%
  3. 3 GPT-5 Chat 76.6%
  4. 4 Doubao-V-Pro 70.1%
  5. 5 Qwen2.5-VL-72B 68.4%
  6. 6 Claude Sonnet 4 66.9%
  7. 7 Qwen2.5-VL-7B 49.3%
  8. 8 UI-TARS-1.5 48.6%
  9. 9 GLM-4.5 44.6%

Source: Official GUI Knowledge Bench leaderboard

#EntryScoreSettingDateSource
1 o3 78.9% Apr 2025 Official leaderboard
2 Gemini 2.5 Pro Jun 2025 77.1% Jun 2025 Official leaderboard
3 GPT-5 Chat 76.6% Aug 2025 Official leaderboard
4 Doubao-V-Pro 70.1% Official leaderboard
5 Qwen2.5-VL-72B 68.4% Jan 2025 Official leaderboard
6 Claude Sonnet 4 66.9% May 2025 Official leaderboard
7 Qwen2.5-VL-7B 49.3% Jan 2025 Official leaderboard
8 UI-TARS-1.5 48.6% Apr 2025 Official leaderboard
9 GLM-4.5 44.6% Aug 2025 Official leaderboard
#ModelScoreSettingDateSource
1 o3 78.9% Apr 2025 Official leaderboard
2 Gemini 2.5 Pro Jun 2025 77.1% Jun 2025 Official leaderboard
3 GPT-5 Chat 76.6% Aug 2025 Official leaderboard
4 Qwen2.5-VL-72B 68.4% Jan 2025 Official leaderboard
5 Claude Sonnet 4 66.9% May 2025 Official leaderboard
6 Qwen2.5-VL-7B 49.3% Jan 2025 Official leaderboard
7 UI-TARS-1.5 48.6% Apr 2025 Official leaderboard
8 GLM-4.5 44.6% Aug 2025 Official leaderboard

Rankings by part

We read 8 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as State, Widget and Layout. o3 leads 7 of them, and Gemini 2.5 Pro leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

State

8 models we track have a score here.

  1. 1 o3 83%
  2. 2 Gemini 2.5 Pro Jun 2025 81.2%
  3. 3 GPT-5 Chat 78.9%
  4. 4 Claude Sonnet 4 ↑1 70.2%
  5. 5 Qwen2.5-VL-72B ↓1 69.7%
  6. 6 Qwen2.5-VL-7B 53.2%
  7. 7 UI-TARS-1.5 49.5%
  8. 8 GLM-4.5 49.5%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Own leaderboard Not tested since Oct 2025
o3 78.9%

Source: Official GUI Knowledge Bench leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2025 Near the ceiling (85%) o3 Qwen2.5-VL-72B 0% 20% 40% 60% 80% 100% 2025 Near the ceiling (85%) o3 Qwen2.5-VL-72B

Source: Official GUI Knowledge Bench 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

All 119 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 15 months ago, in August 2025, for GPT-5 Chat. 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 (9 entries).
  • Pass Models scored 8 models we track, plus 1 other entry on its leaderboard.
  • Warning Newest score The newest came 15 months ago, in August 2025, for GPT-5 Chat.
  • Note Frontier models None of the top 10 has a score. The strongest tested is o3 (#84).
  • Pass Headroom The best is 78.9%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on computer-use-agents.github.io: 8 rows for models we track and 1 other entry 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.

How we check