Benchmark Almanac

UniGUI-Bench

Seed-1.8-VL 84.2% tops its leaderboard Kimi K2.5 80.9% leads the models we track
Not in model cards

Agents interpret and operate graphical interfaces. The benchmark tests 11 capabilities and includes separate system evaluations.

Released
May 2026
Kind
Finding things on screen
Level
Everyday
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English, Chinese
In and out
Text, Screens → Actions
Interaction
Agentic
Refresh
One fixed set
Run by
Anyone
In views
Computer use, Agents
Strongest model tested
Gemini 3 Flash #60 on the capability index
Last model added
April 2026
Scores refreshed
November 20th, 2026
14
Entries in its paper, 11 models we track

Seed-1.8-VL tops the leaderboard with 84.2% Kimi K2.5 leads with 80.9% among models released by April 2026

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. 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. The best model we track is Kimi K2.5, at 80.9%. This view shows only models we track, each at its best.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Seed-1.8-VL 84.2%
  2. 2 Kimi K2.5 80.9%
  3. 3 Qwen3.5-35B-A3B 78.3%
  4. 4 GLM-5V-Turbo 78%
  5. 5 Qwen3-VL-32B 76%
  6. 6 Gemini 3 Flash 74.9%
  7. 7 Seed1.5-VL 71.7%
  8. 8 GLM-4.6V-Flash 66.3%
  9. 9 Gemini 2.5 Flash Sep 2025 65.2%
  10. 10 Qwen3-VL-8B 61.9%
  11. 11 InternVL3-38B 49.2%
  12. 12 InternVL3.5-8B 43.2%
  13. 13 InternVL3.5-30B-A3B 42.5%
  14. 14 InternVL3-8B 35.6%

Source: The UniGUI-Bench paper

#EntryScoreSettingDateSource
1 Seed-1.8-VL 84.2% Official leaderboard
2 Kimi K2.5 80.9% Jan 2026 Official leaderboard
3 Qwen3.5-35B-A3B 78.3% Feb 2026 Official leaderboard
4 GLM-5V-Turbo 78% Apr 2026 Official leaderboard
5 Qwen3-VL-32B 76% Official leaderboard
6 Gemini 3 Flash 74.9% Dec 2025 Official leaderboard
7 Seed1.5-VL 71.7% May 2025 Official leaderboard
8 GLM-4.6V-Flash 66.3% Dec 2025 Official leaderboard
9 Gemini 2.5 Flash Sep 2025 65.2% Sep 2025 Official leaderboard
10 Qwen3-VL-8B 61.9% Official leaderboard
11 InternVL3-38B 49.2% Apr 2025 Official leaderboard
12 InternVL3.5-8B 43.2% Aug 2025 Official leaderboard
13 InternVL3.5-30B-A3B 42.5% Aug 2025 Official leaderboard
14 InternVL3-8B 35.6% Apr 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Kimi K2.5 80.9% Jan 2026 Official leaderboard
2 Qwen3.5-35B-A3B 78.3% Feb 2026 Official leaderboard
3 GLM-5V-Turbo 78% Apr 2026 Official leaderboard
4 Gemini 3 Flash 74.9% Dec 2025 Official leaderboard
5 Seed1.5-VL 71.7% May 2025 Official leaderboard
6 GLM-4.6V-Flash 66.3% Dec 2025 Official leaderboard
7 Gemini 2.5 Flash Sep 2025 65.2% Sep 2025 Official leaderboard
8 InternVL3-38B 49.2% Apr 2025 Official leaderboard
9 InternVL3.5-8B 43.2% Aug 2025 Official leaderboard
10 InternVL3.5-30B-A3B 42.5% Aug 2025 Official leaderboard
11 InternVL3-8B 35.6% Apr 2025 Official leaderboard

Rankings by part

We read 11 parts from the benchmark’s paper, each ranking models on one part of the test, such as Element Grounding, Action Understanding and Page Understanding. Kimi K2.5 leads 5 of them, and 3 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Element Grounding

11 models we track have a score here.

  1. 1 GLM-5V-Turbo ↑2 94%
  2. 2 Gemini 3 Flash ↑2 91%
  3. 3 GLM-4.6V-Flash ↑3 90%
  4. 4 Qwen3.5-35B-A3B ↓2 88.5%
  5. 5 Kimi K2.5 ↓4 85%
  6. 6 Seed1.5-VL ↓1 81.5%
  7. 7 InternVL3-38B ↑1 62%
  8. 8 InternVL3.5-8B ↑1 44%
  9. 9 Gemini 2.5 Flash Sep 2025 ↓2 44%
  10. 10 InternVL3.5-30B-A3B 34%

Number one over time

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

Moonshot AI
Benchmark paper Not tested since May 2026
Kimi K2.5 80.9%

Source: The UniGUI-Bench paper

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
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Kimi K2.5 Gemini 3 Flash Seed1.5-VL InternVL3-38B 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Kimi K2.5 Seed1.5-VL

Source: Official UniGUI-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 7 months ago, in April 2026, for GLM-5V-Turbo. 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 benchmark’s paper (14 entries).
  • Pass Models scored 11 models we track, plus 3 other entries on its leaderboard.
  • Warning Newest score The newest came 7 months ago, in April 2026, for GLM-5V-Turbo.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 3 Flash (#60).
  • Pass Headroom The best is 80.9%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on aclanthology.org: 11 rows for models we track and 3 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.

How we check