Mistral Small 2603
Mistral Small 2603 scores 12.5 out of 100 overall, which puts it in a tie for rank 34–36 of 36 models on the AI Vision Index. It is strongest at Not being fooled (22) and weakest at Space, 3D and motion (7). It beats the average of all 36 models in none of the six categories. It cannot take the two video tests, so its score covers the other tests.
Overall score12.595% interval 11.5–14.2
Rank34–36of 36 models (tied)
Cost per 1,000 answers$0.26
Seconds per answer3.0 s
Thinking effortdefault
Maker Mistral
WeightsOpen weights
AccessAPI
Tested
Reading15
Counting10
Space, 3D and motion7
Spotting changes and errors12
Diagrams10
Not being fooled22
The bar is the score; the light band is the 95% interval. The overall score is the average of the six.
Reading
Reading a clock0
Reading a meter6
Eye chart0
Reading a label61
Reading a chart29
Test strips22
Adding up coins2
Traffic light from a car0
Counting
Counting objects6
Counting stacked goods14
Counting on a conveyorNot supported—
Space, 3D and motion
Which is nearer26
Which way it faces0
Same or mirror image0
Will it topple7
Match the render6
Which bay can I park in?0
Follow the hidden objectNot supported—
Spotting changes and errors
Spot the change32
Shelf against plan4
Rendering glitch6
Check the invoice6
Diagrams
Follow the cable10
Follow the pipe2
Missing dimension16
Not being fooled
Absent objects38
Saying "I can't tell"50
Optical illusions0
Real or AI-made0
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
- Gemma 4 E4B 13.7 Compare
- LFM2.5-VL 1.6B 12.4 Compare
Badge for your site
HTML
<a href="https://aivisionindex.com/models/mistral-small-2603/"><img src="https://aivisionindex.com/badge/mistral-small-2603.svg" alt="Mistral Small 2603: ranked #34–36 of 36 on the AI Vision Index"></a>Markdown
[](https://aivisionindex.com/models/mistral-small-2603/)Results are CC BY 4.0, so you are free to reuse them if you credit the AI Vision Index. All results (JSON)