Released: 2026-09-10
mistralai/mistral-medium-3
mistral
3
5
1
2
4
6
8
9
7
2
8
7
3
5
9
6
4
1
4
6
9
7
8
1
3
2
5
1
2
3
4
9
5
8
7
6
8
4
5
6
1
7
2
3
9
9
7
6
8
2
3
5
1
4
7
9
8
9
6
4
4
5
2
6
1
2
1
3
7
9
8
4
5
3
4
5
7
2
1
6
8
openai/gpt-4o
openai
3
5
9
6
4
1
2
7
8
8
1
7
2
5
9
6
4
3
4
6
2
7
8
3
9
1
5
9
2
3
5
1
7
8
6
4
5
4
1
6
3
8
7
2
9
7
8
6
4
2
9
5
3
1
7
3
4
9
6
5
1
8
2
6
9
8
1
7
2
3
5
4
2
7
5
3
9
8
4
6
1
google/gemini-3-flash-preview
3
5
9
2
4
6
1
7
8
2
8
7
5
1
9
6
4
3
4
6
1
7
8
3
9
5
2
5
1
2
4
9
2
8
6
7
8
4
3
6
7
5
2
1
9
9
7
6
3
2
8
5
4
1
7
2
8
9
3
4
4
1
5
6
3
5
1
5
7
4
8
2
1
9
4
8
6
2
7
3
6
meta-llama/llama-3.3-70b-instruct
meta
5
3
4
6
7
8
9
1
2
6
7
2
1
9
5
3
4
8
1
9
8
3
4
2
5
6
7
8
5
9
7
6
1
4
2
3
4
2
6
8
5
3
7
9
1
7
1
3
9
2
4
8
5
6
9
6
1
5
3
7
2
8
4
2
8
7
4
1
9
6
3
5
3
4
5
2
8
6
1
7
9
deepseek/deepseek-chat-v3-0324
deepseek
3
5
8
2
4
6
1
9
7
1
2
7
8
3
9
6
4
5
4
6
9
7
8
5
2
1
3
2
1
5
3
9
4
8
7
6
8
4
3
6
7
1
2
5
9
9
7
6
5
2
8
5
3
1
7
3
2
9
5
4
4
8
6
6
9
4
1
6
7
3
2
8
5
8
1
4
6
2
7
3
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.