Released: 2026-09-13
mistralai/mistral-medium-3
mistral
2
7
1
3
6
4
5
9
8
8
4
9
5
3
6
7
1
2
5
6
3
9
8
2
2
4
7
6
3
7
1
4
9
8
2
5
9
5
4
8
2
3
6
7
1
3
1
2
6
5
7
4
9
8
4
9
8
2
7
5
3
6
1
1
2
6
4
9
8
7
5
3
7
8
5
3
1
6
9
4
2
google/gemini-3-flash-preview
2
3
1
7
9
4
5
8
6
7
8
9
5
3
6
4
1
2
5
4
6
1
8
2
2
7
9
6
7
8
2
4
5
1
9
3
9
5
4
8
2
1
6
3
7
3
1
2
9
6
7
8
5
4
4
6
5
3
7
8
9
2
1
1
2
7
6
4
9
3
8
5
8
9
3
4
5
2
7
6
1
openai/gpt-4o
openai
8
3
1
2
9
4
5
7
6
2
7
9
5
3
6
1
4
8
5
6
4
7
8
1
2
3
9
6
2
7
1
4
5
8
9
3
9
5
8
8
2
3
6
1
7
3
4
2
6
8
7
9
5
1
4
9
5
3
7
8
6
2
1
1
3
6
9
1
2
7
8
5
7
8
3
4
6
9
3
6
2
meta-llama/llama-3.3-70b-instruct
meta
5
3
1
9
6
4
8
2
7
2
8
6
5
3
7
1
9
4
4
9
7
1
8
2
6
5
3
6
1
5
7
4
3
9
8
2
8
5
4
2
9
1
7
6
3
3
7
2
8
5
6
4
1
9
9
4
8
6
7
5
3
2
1
1
2
3
4
6
9
5
7
8
7
6
9
3
1
8
2
4
5
deepseek/deepseek-chat-v3-0324
deepseek
8
9
1
2
6
4
5
7
3
2
4
7
5
3
6
8
1
9
5
3
6
7
8
9
2
4
1
6
7
9
1
4
8
3
5
2
4
5
1
8
2
3
6
9
7
3
8
2
6
9
7
1
5
4
4
6
8
9
7
5
1
2
3
1
2
4
3
6
1
9
8
5
7
5
3
4
1
2
8
6
9
Each model receives the same Sudoku puzzle with a partially filled grid. Models are tasked with solving the puzzle by filling in all empty cells following Sudoku rules, and their solutions are evaluated without feedback or retries.