Released: 2026-07-30
openai/gpt-4o
openai
4
7
2
1
9
8
6
5
3
3
9
5
2
6
4
8
7
1
6
1
8
5
3
7
2
4
9
2
5
6
8
4
3
1
9
7
1
8
7
9
5
6
4
3
2
9
3
4
7
2
1
5
8
6
5
1
3
6
8
2
9
7
4
8
6
2
3
7
9
7
1
5
7
4
9
1
7
5
3
2
8
google/gemini-3-flash-preview
7
9
2
1
8
4
6
5
3
3
4
5
2
6
9
8
7
1
6
1
8
5
3
7
2
4
9
1
5
6
8
4
3
2
9
7
2
7
3
9
5
6
4
1
8
9
8
4
4
7
2
3
6
5
5
1
7
6
2
8
9
3
4
8
2
9
3
7
1
5
6
0
4
3
9
7
1
5
6
2
8
mistralai/mistral-medium-3
mistral
1
4
7
9
2
8
6
5
3
3
9
5
4
6
1
8
7
2
6
2
8
5
3
7
2
1
4
2
5
6
8
4
3
5
9
7
7
3
9
2
5
6
4
8
1
4
8
1
9
7
5
3
2
6
5
1
2
6
8
4
7
3
9
9
6
4
3
1
2
9
7
5
8
7
3
7
9
9
1
4
8
meta-llama/llama-3.3-70b-instruct
meta
4
8
5
3
1
2
6
5
3
3
1
2
9
6
8
7
4
5
6
9
8
5
4
7
2
1
3
9
5
3
8
4
6
1
2
7
1
4
6
2
5
3
8
9
7
2
7
1
4
7
5
9
6
8
5
6
4
1
3
9
7
2
8
8
3
9
7
2
1
4
5
6
7
2
6
4
8
5
3
1
9
deepseek/deepseek-chat-v3-0324
deepseek
9
7
4
1
2
8
6
5
3
3
2
1
4
6
5
8
7
9
6
5
8
9
3
7
2
1
4
2
5
6
8
4
3
1
9
7
7
3
9
2
1
6
4
8
5
1
4
8
5
7
9
3
6
2
5
1
3
6
8
2
7
4
9
8
6
7
3
9
1
5
2
4
4
9
2
7
5
4
6
3
8
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.