GLM-5.3 Flash
GLM-5.3 Flash scores 43.2 out of 100 overall, which puts it in a tie for rank 19–29 of 36 models on the AI Vision Index. It is strongest at Not being fooled (57) and weakest at Counting (16). It beats the average of all 36 models in none of the six categories.
Overall score43.295% interval 41.7–44.7
Rank19–29of 36 models (tied)
Cost per 1,000 answers$1.40
Seconds per answer9.8 s
Thinking effortdefault
Maker Zhipu
WeightsOpen weights
AccessAPI
Tested
Reading42
Counting16
Space, 3D and motion39
Spotting changes and errors56
Diagrams49
Not being fooled57
The bar is the score; the light band is the 95% interval. The overall score is the average of the six.
Reading
Reading a clock42
Reading a meter43
Eye chart26
Reading a label61
Reading a chart44
Test strips52
Adding up coins25
Traffic light from a car46
Counting
Counting objects26
Counting stacked goods11
Counting on a conveyor12
Space, 3D and motion
Which is nearer54
Which way it faces45
Same or mirror image0
Will it topple18
Match the render53
Which bay can I park in?75
Follow the hidden object30
Spotting changes and errors
Spot the change56
Shelf against plan58
Rendering glitch27
Check the invoice82
Diagrams
Follow the cable40
Follow the pipe37
Missing dimension70
Not being fooled
Absent objects89
Saying "I can't tell"90
Optical illusions11
Real or AI-made36
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
- DeepSeek V4.1 Flash 44.4 Compare
- Qwen3.6-35B-A3B 43.0 Compare
Badge for your site
HTML
<a href="https://aivisionindex.com/models/glm-5-3-flash/"><img src="https://aivisionindex.com/badge/glm-5-3-flash.svg" alt="GLM-5.3 Flash: ranked #19–29 of 36 on the AI Vision Index"></a>Markdown
[](https://aivisionindex.com/models/glm-5-3-flash/)Results are CC BY 4.0, so you are free to reuse them if you credit the AI Vision Index. All results (JSON)