Gemini 3.8 Flash
Gemini 3.8 Flash scores 61.9 out of 100 overall, which puts it in a tie for rank 10–18 of 36 models on the AI Vision Index. It is strongest at Spotting changes and errors (85) and weakest at Counting (42). It beats the average of all 36 models in five of the six categories.
Overall score61.995% interval 60.2–63.7
Rank10–18of 36 models (tied)
Cost per 1,000 answers$12.80
Seconds per answer11.5 s
Thinking effortmedium
Maker Google
WeightsClosed weights
AccessAPI
Tested
Reading58
Counting42
Space, 3D and motion66
Spotting changes and errors85
Diagrams68
Not being fooled53
The bar is the score; the light band is the 95% interval. The overall score is the average of the six.
Reading
Reading a clock72
Reading a meter51
Eye chart48
Reading a label82
Reading a chart48
Test strips38
Adding up coins53
Traffic light from a car75
Counting
Counting objects30
Counting stacked goods54
Counting on a conveyor41
Space, 3D and motion
Which is nearer75
Which way it faces76
Same or mirror image26
Will it topple62
Match the render72
Which bay can I park in?93
Follow the hidden object56
Spotting changes and errors
Spot the change80
Shelf against plan92
Rendering glitch67
Check the invoice99
Diagrams
Follow the cable58
Follow the pipe65
Missing dimension81
Not being fooled
Absent objects95
Saying "I can't tell"97
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
- Claude Haiku 5.5 63.5 Compare
- GPT-6 Luna 60.7 Compare
Badge for your site
HTML
<a href="https://aivisionindex.com/models/gemini-3-8-flash/"><img src="https://aivisionindex.com/badge/gemini-3-8-flash.svg" alt="Gemini 3.8 Flash: ranked #10–18 of 36 on the AI Vision Index"></a>Markdown
[](https://aivisionindex.com/models/gemini-3-8-flash/)Results are CC BY 4.0, so you are free to reuse them if you credit the AI Vision Index. All results (JSON)