Skip to main content
BenchLM
Data

Scale Labs Visual-Language Understanding (Visual-Language Understanding)

We show this table for reference; we do not rank on it.

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Scale score on Visual-Language Understanding — November 12, 2026 snapshot

We mirror the published scale score view for Visual-Language Understanding. Gemini 2.5 Pro Experimental (March 2025) leads the public snapshot at 54.6%, followed by gemini-2.5-pro-preview-06-05 (54.6%) and GPT-5.4 Pro (53.9%). We do not use these results to rank models overall.

1

Gemini 2.5 Pro Experimental (March 2025)

Google

54.6%
2

gemini-2.5-pro-preview-06-05

Google

54.6%
3Closed

GPT-5.4 Pro

OpenAI

53.9%
Overall 66.96Context 1.05M

63 modelsMultimodal & GroundedCurrentDisplay onlyUpdated November 12, 2026 snapshot

Scale score table (63 models)

Scale score results for Visual-Language Understanding
RankModel / configurationScoreParameters (B)Open / closed
1Gemini 2.5 Pro Experimental (March 2025)Google
54.6%
Not reportedUnknown
2gemini-2.5-pro-preview-06-05Google
54.6%
Not reportedUnknown
3GPT-5.4 ProOpenAI
53.9%
Not reportedClosed
4gpt-5-pro-2025-10-06OpenAI
52.4%
Not reportedUnknown
5o4-mini (high) (April 2025)OpenAI
51.8%
Not reportedUnknown
6o4-mini (medium) (April 2025)OpenAI
51.7%
Not reportedUnknown
7o3 Pro (high) (June 2025)OpenAI
51.6%
Not reportedUnknown
8gemini-3-pro-previewGoogle
51.5%
Not reportedUnknown
9gpt-5.4-2026-03-05 (xhigh thinking)OpenAI
50.9%
Not reportedUnknown
10Gemini 2.5 Pro Preview (May 06 2025)Google
50.8%
Not reportedUnknown
11GPT-5 miniOpenAI
50.4%
Not reportedClosed
12o3 (high) (April 2025)OpenAI
50.1%
Not reportedUnknown
13gpt-5-2025-08-07OpenAI
49.7%
Not reportedUnknown
14o3 (medium) (April 2025)OpenAI
49.6%
Not reportedUnknown
15Gemini 2.5 Flash Preview (May 2025)Google
49.1%
Not reportedUnknown
16Claude Sonnet 4.5 ThinkingAnthropic
48.8%
Not reportedClosed
17claude-opus-4-1-20250805-thinkingAnthropic
48.4%
Not reportedUnknown
18Claude 3.7 Sonnet Thinking (Feb 2025)Anthropic
48.2%
Not reportedUnknown
19o1 Pro (March 2025)OpenAI
47.3%
Not reportedUnknown
20Gemini 2.5 Flash (April 2025)Google
47.0%
Not reportedUnknown
21Claude Opus 4 (Thinking)Anthropic
47.0%
Not reportedUnknown
22gemini-3.1-flash-lite-previewGoogle
46.9%
Not reportedUnknown
23GPT-5.2OpenAI
46.6%
Not reportedClosed
24Claude Opus 4.5 ThinkingAnthropic
46.4%
Not reportedClosed
25claude-opus-4-6-thinking-maxAnthropic
46.1%
Not reportedUnknown
26Gemini 2.0 Flash Thinking ExperimentalGoogle
45.5%
Not reportedUnknown
27Claude Sonnet 4 (Thinking)Anthropic
45.5%
Not reportedUnknown
28claude-opus-4-6 (Non-Thinking)Unknown
45.5%
Not reportedUnknown
29GPT-4.1OpenAI
45.3%
Not reportedClosed
30Claude Opus 4.5Anthropic
45.3%
Not reportedClosed
31o1 (December 2024)OpenAI
45.3%
Not reportedUnknown
32claude-opus-4-1-20250805Anthropic
45.3%
Not reportedUnknown
33Claude Sonnet 4.5Anthropic
45%
Not reportedClosed
34gpt-5.1-thinkingOpenAI
43.8%
Not reportedUnknown
35Claude Opus 4Anthropic
43.5%
Not reportedUnknown
36Gemini 2.0 Pro Experimental (Feb 2025)Google
43.3%
Not reportedUnknown
37Claude Sonnet 4Anthropic
43.2%
Not reportedUnknown
38Claude 3.7 Sonnet (February 2025)Anthropic
43.0%
Not reportedUnknown
39GPT-4.5 Preview (February 2025)OpenAI
42.1%
Not reportedUnknown
40Kimi K2.5Moonshot AI
41.9%
Not reportedOpen
41GPT-4.1 miniOpenAI
41.1%
Not reportedClosed
42Gemini 2.0 Flash ExperimentalGoogle
40.0%
Not reportedUnknown
43Gemini 2.0 Flash (February 2025)Google
39.9%
Not reportedUnknown
44Claude 3.5 Sonnet (October 2024)Anthropic
38.7%
Not reportedUnknown
45Claude 3.5 Sonnet (June 2024)Anthropic
38.4%
Not reportedUnknown
46Llama 4 MaverickMeta
38.3%
Not reportedOpen
47ChatGPT-4o-latest (November 2024)OpenAI
38.0%
Not reportedUnknown
48Gemini 1.5 ProGoogle
37.1%
Not reportedClosed
49GPT-4o (August 2024)OpenAI
34.9%
Not reportedUnknown
50gpt-5.1-instantOpenAI
34.9%
Not reportedUnknown
51Mistral Medium 3Mistral
34.6%
Not reportedClosed
52Gemini 1.5 Flash 002Google
34.0%
Not reportedUnknown
53Pixtral Large (November 2024)Mistral
33.9%
Not reportedUnknown
54Gemini 2.0 Flash Lite PreviewGoogle
32.7%
Not reportedUnknown
55Qwen2-VL-72B-InstructAlibaba
28.6%
Not reportedUnknown
56Claude 3 OpusAnthropic
27.8%
Not reportedClosed
57GPT-4.1 nanoOpenAI
26.6%
Not reportedClosed
58Nova ProAmazon
26.3%
Not reportedClosed
59Pixtral 12B (September 2024)Mistral
26.0%
Not reportedUnknown
60Nova LiteAmazon
25.5%
Not reportedUnknown
61Llama 3.2 90B Vision InstructMeta
24.6%
Not reportedUnknown
62Llama 3.2 11B Vision-InstructMeta
20.5%
Not reportedUnknown
63Phi 3.5 Vision-InstructMicrosoft
15.2%
Not reportedUnknown

How to read this leaderboard

Compare the published configurations as complete evaluation systems. The source can combine a base model, agent scaffold, tools, budget, and inference setting in each result.

Operator receipt: 63 sourced rows are currently displayable on this page; the leading published row is Gemini 2.5 Pro Experimental (March 2025) at 54.6%.

Honest limit: This Scale table is display-only context, not benchmark provenance or a weighted model-only comparison.

Visual-Language Understanding on Scale Labs

How BenchLM shows Visual-Language Understanding

BenchLM mirrors 63 published rows from Scale Labs’ public Visual-Language Understanding leaderboard, captured on November 12, 2026 snapshot.

The table is display only. It is useful context for a published agent or model configuration, but it does not enter BenchLM’s overall or category rankings. Scale has not added a row to this board since March 2026.

Snapshot

63 published rowsScale Labs sourceNo new rowsDisplay only
Visual-Language Understanding on Scale Labs

The published Visual-Language Understanding snapshot places Gemini 2.5 Pro Experimental (March 2025) first at 54.6%. The third row is 0.8 points behind. The broader top-10 range is 3.9 points, so many of the published results sit in a relatively narrow band.

63 models have been evaluated on Visual-Language Understanding. The benchmark falls in the Multimodal & Grounded category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. Visual-Language Understanding is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About Visual-Language Understanding

Year

2026

Tasks

63 published rows

Format

Published Scale leaderboard score

Difficulty

External agent and model evaluation

BenchLM mirrors 63 published rows from the Visual-Language Understanding public table captured on November 12, 2026 snapshot.

Scale Labs leaderboardPublic benchmark source

Freshness and provenance

Version

Visual-Language Understanding 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does Visual-Language Understanding measure?

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Which model leads the published Visual-Language Understanding snapshot?

Gemini 2.5 Pro Experimental (March 2025) currently leads the published Visual-Language Understanding snapshot with 54.6% scale score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on Visual-Language Understanding?

The November 12, 2026 snapshot snapshot contains 63 AI models.

Last updated: November 12, 2026 snapshot · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 7,000+ subscribers.

One email each week. Unsubscribe anytime.