Released: 2026-06-19
openai/gpt-4o
openai
4
8
7
2
1
5
9
3
6
1
3
2
8
7
6
5
4
9
9
5
6
3
4
1
2
7
8
8
1
3
6
2
4
7
9
5
7
2
4
5
9
8
6
1
3
5
6
9
1
3
7
4
8
2
3
9
8
4
5
2
1
6
7
6
7
1
9
8
3
3
2
4
2
4
5
7
6
9
8
5
1
google/gemini-3-flash-preview
4
8
9
2
7
5
3
6
1
1
3
2
8
4
6
7
5
9
7
5
6
1
3
9
2
8
4
8
1
3
5
2
4
9
7
6
5
6
4
7
9
8
2
1
3
9
2
7
6
3
7
4
5
8
3
9
8
4
5
2
1
6
7
2
7
1
9
8
3
6
4
5
6
4
5
3
1
7
5
9
2
deepseek/deepseek-chat-v3-0324
deepseek
4
8
9
2
1
5
7
6
3
1
3
2
8
7
6
5
4
9
7
5
6
3
4
9
2
8
1
8
1
7
6
5
4
9
3
2
2
6
4
5
9
8
7
1
3
9
5
3
1
3
7
4
2
6
3
9
8
7
5
2
1
6
4
5
7
1
9
8
3
6
2
5
6
2
4
4
6
1
5
9
7
meta-llama/llama-3.3-70b-instruct
meta
4
2
8
2
5
9
6
1
7
3
9
1
4
7
6
5
8
2
5
6
2
1
3
8
4
9
7
8
1
9
5
2
4
3
6
5
2
4
7
9
1
3
8
5
6
6
5
3
8
4
7
2
1
9
1
8
5
6
9
2
7
4
3
9
7
4
3
8
1
5
2
6
7
3
6
5
6
4
1
8
9
mistralai/mistral-medium-3
mistral
4
1
7
2
6
8
3
5
9
2
3
8
4
5
6
7
1
9
9
5
6
3
7
1
2
8
4
8
1
3
5
2
4
9
6
7
5
2
4
6
9
3
8
1
7
6
9
7
1
3
7
4
2
8
3
9
8
7
5
2
1
4
6
1
7
1
9
8
5
6
3
2
7
4
2
8
1
6
5
9
3
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.