Qwen3.5-9B
Qwen3.5-9B scores 29.5 out of 100 overall, which puts it in a tie for rank 30–31 of 36 models on the AI Vision Index. It is strongest at Not being fooled (46) and weakest at Counting (19). It beats the average of all 36 models in none of the six categories.
Overall score29.595% interval 28.0–31.1
Rank30–31of 36 models (tied)
Cost per 1,000 answers$0 (local)
Seconds per answer1.5 s on a laptop GPU
Thinking effortno thinking
Maker Alibaba
WeightsOpen weights
AccessRuns locally
Tested
Reading36
Counting19
Space, 3D and motion27
Spotting changes and errors24
Diagrams25
Not being fooled46
The bar is the score; the light band is the 95% interval. The overall score is the average of the six.
Reading
Reading a clock50
Reading a meter38
Eye chart3
Reading a label68
Reading a chart47
Test strips33
Adding up coins4
Traffic light from a car50
Counting
Counting objects19
Counting stacked goods10
Counting on a conveyor28
Space, 3D and motion
Which is nearer60
Which way it faces23
Same or mirror image13
Will it topple10
Match the render50
Which bay can I park in?31
Follow the hidden object0
Spotting changes and errors
Spot the change48
Shelf against plan8
Rendering glitch31
Check the invoice8
Diagrams
Follow the cable35
Follow the pipe6
Missing dimension34
Not being fooled
Absent objects89
Saying "I can't tell"73
Optical illusions0
Real or AI-made20
Scores are corrected for guessing: 0 is no better than a guess, 100 is always right. All models on every test
Neighbours on the leaderboard
- Hunyuan HY4 37.9 Compare
- Perceptron Mk1.5 28.8 Compare
Badge for your site
HTML
<a href="https://aivisionindex.com/models/qwen3-5-9b/"><img src="https://aivisionindex.com/badge/qwen3-5-9b.svg" alt="Qwen3.5-9B: ranked #30–31 of 36 on the AI Vision Index"></a>Markdown
[](https://aivisionindex.com/models/qwen3-5-9b/)Results are CC BY 4.0, so you are free to reuse them if you credit the AI Vision Index. All results (JSON)