Benchmark Almanac

VisualProbe

Fading

Models search high-resolution images for small targets to answer questions, often over many crop and zoom turns. Answers are checked against a key.

Released with the Mini-o3 paper as the Visual Probe Dataset, with train and test splits.

Released
September 2025
Kind
Detection and captioning
Level
Everyday
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Last in a launch post
May 2026, dropped before it saturated
Last model added
May 2026
1
Model scored in lab cards
1
Launch post lists it, from 1 lab

Step 3.7 Flash leads with 65%

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

  1. 1 Step 3.7 Flash 65%

Source: Labs’ model cards

#ModelScoreSettingDateSource
1 Step 3.7 Flash 65% May 2026 Lab model card

Scores over time

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

Reported by the lab in its model card
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Step 3.7 Flash 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Step 3.7 Flash

Source: Labs’ model cards

Labs headline it less often now, last in May 2026

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

StepFun
DateLaunch postScoreAs the post writes it
May 29, 2026 Step 3.7 Flash 65 VisualProbe (with Python)

More in Vision

All 281 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. GPT-5.1 85.4%
  • MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
  • CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
  • MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
  • MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%
  • ChartQA In use Models answer human-written and generated questions about chart images that need visual and arithmetic reasoning. Scored by relaxed accuracy with a 5 percent numeric tolerance. Command A Vision 90.9%