Released: 2026-05-28
openai/gpt-4o
openai
2
4
3
1
6
8
5
7
9
6
8
1
9
5
7
3
4
2
5
9
7
2
3
4
8
1
6
8
5
2
6
7
3
4
9
1
3
6
9
5
8
1
7
2
4
7
1
4
8
9
2
6
3
5
9
3
6
7
4
5
1
8
2
4
2
5
3
1
9
9
5
7
1
7
8
4
2
6
9
6
3
google/gemini-2.0-flash-001
3
6
4
9
1
2
5
7
8
5
8
1
6
4
7
3
9
2
9
2
7
5
3
4
8
1
6
8
4
5
1
7
3
9
6
2
6
9
3
4
5
8
7
2
1
7
1
2
8
9
6
4
3
5
2
3
6
7
4
5
1
8
9
4
7
9
3
8
1
2
5
6
1
5
8
2
6
9
6
4
3
google/gemini-3-flash-preview
3
4
1
2
6
8
5
7
9
9
8
5
1
0
7
4
6
2
2
6
7
9
3
4
8
1
5
5
2
4
6
1
3
7
9
8
6
7
9
5
8
4
3
2
1
8
1
2
7
9
5
6
3
4
5
3
6
8
4
7
1
9
2
4
2
9
3
6
1
8
5
7
1
5
8
4
2
9
3
7
6
mistralai/mistral-medium-3
mistral
1
4
3
6
8
9
5
7
2
5
8
9
1
6
7
3
4
2
6
2
7
5
3
4
8
9
1
7
5
4
8
1
3
2
6
9
8
6
2
9
7
5
4
3
1
3
1
9
2
4
6
7
8
5
2
3
6
7
4
8
1
5
9
4
7
1
3
5
9
6
2
8
9
5
8
4
2
1
7
3
6
meta-llama/llama-3.3-70b-instruct
meta
5
9
1
7
6
8
5
4
3
8
2
3
1
5
7
9
6
2
4
6
7
9
3
4
2
8
1
9
5
2
8
1
3
7
1
6
1
4
8
6
9
5
3
2
1
6
1
5
4
9
2
8
3
5
3
7
9
5
8
1
1
9
4
2
8
6
3
7
9
5
1
7
7
3
4
2
6
1
8
5
9
deepseek/deepseek-chat-v3-0324
deepseek
1
2
3
6
8
9
5
7
4
5
8
4
1
6
7
3
9
2
6
9
7
2
3
4
8
1
5
2
6
5
4
1
3
7
8
9
3
4
9
7
5
8
2
6
1
7
1
8
9
2
6
4
3
5
8
3
6
5
4
2
1
9
7
4
7
2
8
9
1
6
5
3
9
5
1
3
7
6
4
2
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.