Released: 2026-06-11
openai/gpt-4o
openai
9
3
8
6
7
1
5
4
2
4
1
2
5
8
3
7
9
6
7
5
6
4
9
2
8
3
1
6
9
1
8
4
7
0
5
3
8
2
9
3
1
5
4
7
6
3
7
4
2
6
9
0
1
8
1
6
7
9
5
4
3
2
8
5
8
2
1
3
6
9
0
7
2
4
3
7
0
8
0
6
9
google/gemini-3-flash-preview
3
9
8
4
2
1
5
7
6
4
1
6
5
8
3
7
9
2
7
5
2
6
9
7
8
3
1
6
3
2
8
4
5
9
1
3
8
2
9
3
1
7
4
5
6
5
7
4
9
6
1
2
8
3
1
6
7
2
5
9
3
4
8
2
8
5
1
3
6
0
0
0
9
4
3
7
8
2
1
6
5
meta-llama/llama-3.3-70b-instruct
meta
5
1
9
7
3
8
6
4
2
4
1
6
5
8
2
7
3
9
3
5
2
9
1
4
8
6
7
6
8
7
3
4
5
2
9
1
8
4
5
2
6
9
1
7
3
2
7
1
8
5
6
4
3
9
9
3
4
1
7
2
5
8
6
1
6
8
4
9
7
3
5
2
7
9
3
6
2
1
4
5
8
mistralai/mistral-medium-3
mistral
3
2
6
4
8
1
9
5
7
4
1
8
5
2
9
7
3
6
7
5
9
3
6
7
8
2
1
6
9
2
8
4
5
1
7
3
8
3
9
2
1
7
4
6
5
1
7
5
9
3
4
2
8
6
2
4
1
6
5
3
3
9
8
5
6
7
1
9
8
6
4
2
9
8
3
7
4
2
5
6
1
deepseek/deepseek-chat-v3-0324
deepseek
2
3
7
8
6
4
5
9
1
4
1
8
5
2
9
7
3
6
9
5
6
3
7
1
8
4
2
6
2
1
9
4
7
6
8
3
8
4
9
6
1
3
4
5
7
3
7
5
2
8
6
2
1
9
1
6
2
4
3
5
9
7
8
7
9
4
1
5
8
3
2
4
5
8
3
7
9
2
1
6
5
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.