Released: 2026-05-31
mistralai/mistral-medium-3
mistral
4
5
2
8
6
1
3
9
7
3
9
7
2
5
4
6
1
8
8
6
1
6
1
9
4
7
2
2
1
3
4
7
9
5
8
6
5
7
6
3
4
8
1
2
9
6
4
9
1
3
2
7
6
8
1
2
8
9
7
6
2
3
4
7
3
5
2
8
4
9
6
1
9
8
4
7
2
3
8
5
6
google/gemini-3-flash-preview
6
1
2
9
5
4
3
1
7
4
9
7
2
8
3
6
5
1
3
5
8
6
1
7
4
9
2
1
4
3
5
2
9
5
8
7
2
5
6
1
4
8
7
3
9
8
7
9
4
3
1
1
6
5
9
3
1
5
6
4
2
7
8
7
6
5
3
9
2
8
1
4
5
8
4
7
2
3
9
1
6
google/gemini-2.0-flash-001
4
6
1
9
8
5
3
2
7
3
9
7
2
6
4
8
5
1
8
5
2
6
1
7
4
9
3
2
4
3
1
7
9
5
8
6
5
7
6
8
4
2
1
3
9
1
8
9
5
3
6
7
4
2
6
3
5
4
9
1
2
7
8
7
1
2
3
5
8
6
9
4
9
8
4
7
2
3
5
1
6
openai/gpt-4o
openai
5
6
1
4
8
2
3
9
7
8
9
4
2
7
3
6
1
5
3
7
2
6
1
9
4
5
8
4
1
3
7
6
9
5
8
2
2
5
6
8
4
1
7
3
9
7
8
9
5
3
6
1
2
4
1
4
8
9
5
7
2
6
3
7
3
5
1
2
8
9
4
6
9
8
4
7
6
5
3
2
6
meta-llama/llama-3.3-70b-instruct
meta
5
1
9
7
3
8
6
4
2
2
9
8
1
5
4
3
7
6
4
7
3
9
2
6
1
5
8
8
5
1
4
7
2
9
3
6
6
3
2
8
9
5
7
1
4
9
4
7
6
1
3
5
2
8
1
8
5
3
6
7
2
9
4
7
2
4
5
8
9
6
3
1
3
6
9
2
4
1
8
5
7
deepseek/deepseek-chat-v3-0324
deepseek
5
6
1
8
9
4
3
2
7
3
9
7
2
5
8
6
1
4
8
4
2
6
1
3
4
9
5
4
2
3
1
6
9
5
8
2
1
7
6
5
4
2
8
3
9
2
5
9
7
3
8
1
6
4
9
3
5
4
8
6
2
7
1
7
1
6
9
2
5
4
4
8
2
8
4
7
9
1
5
3
6
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.