Released: 2026-09-08
openai/gpt-4o
openai
1
7
4
9
8
5
2
3
6
9
8
5
3
7
6
1
2
4
6
2
3
1
4
2
5
9
7
8
9
2
7
5
3
4
1
6
5
4
7
6
1
8
7
5
2
3
1
6
8
9
4
7
8
5
4
3
8
5
2
1
6
7
9
7
6
1
4
3
9
8
2
3
2
5
3
2
6
7
9
6
8
mistralai/mistral-medium-3
mistral
3
1
4
9
8
5
2
6
7
7
8
9
3
2
6
1
4
5
6
2
5
4
7
1
3
9
8
8
9
2
5
3
7
4
1
6
1
4
6
2
5
8
9
3
2
5
3
7
8
9
4
6
2
1
4
6
3
7
2
1
8
5
9
2
7
1
6
3
9
5
8
4
9
5
8
1
4
2
7
6
3
google/gemini-3-flash-preview
7
3
1
9
8
5
6
2
4
9
8
4
2
7
6
1
5
3
6
2
5
1
4
3
8
9
7
8
9
2
3
5
7
4
1
6
5
4
7
6
1
8
9
3
2
3
1
6
8
9
4
7
2
5
4
6
9
5
2
1
3
7
8
2
7
8
4
3
9
5
1
6
1
5
3
7
6
2
4
8
9
meta-llama/llama-3.3-70b-instruct
meta
5
3
4
9
8
5
6
7
2
8
6
7
4
1
3
5
2
9
6
2
9
7
5
4
3
8
1
8
5
6
1
4
7
2
9
3
4
1
3
2
9
8
7
5
6
7
9
2
8
6
5
1
4
3
9
7
1
3
2
4
8
6
5
2
8
5
6
3
9
4
1
7
3
4
8
5
7
2
9
6
1
deepseek/deepseek-chat-v3-0324
deepseek
4
7
1
9
8
5
6
2
3
5
8
9
2
6
3
1
4
7
6
2
3
1
4
7
5
9
8
8
9
5
3
2
7
4
1
6
3
4
7
5
1
6
9
8
2
1
6
2
8
9
4
7
3
5
9
3
6
4
2
1
8
7
5
2
1
8
6
3
9
7
5
4
7
5
4
0
0
8
2
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.