Released: 2026-06-26
openai/gpt-4o
openai
8
3
2
6
7
1
5
4
9
4
7
1
5
2
8
9
6
3
6
5
9
4
3
2
7
1
8
9
4
1
2
8
6
7
3
5
5
6
8
9
1
4
2
7
3
7
2
3
7
5
9
1
8
4
2
9
6
1
4
7
3
5
8
3
8
4
9
3
5
6
2
7
1
5
7
8
6
9
4
9
6
mistralai/mistral-medium-3
mistral
5
3
9
6
7
8
4
2
1
4
7
6
5
2
1
9
8
3
6
1
2
9
4
3
7
5
8
9
4
1
2
8
6
7
3
5
3
2
8
7
1
4
5
6
9
7
6
5
3
9
5
1
4
2
2
9
3
8
4
7
6
5
1
8
5
4
1
3
2
6
9
7
1
8
7
6
5
9
2
3
4
google/gemini-3-flash-preview
8
3
5
4
9
2
4
1
6
4
7
2
5
6
8
9
1
3
6
1
9
3
7
5
2
4
8
5
4
1
9
8
6
7
3
2
3
6
8
2
1
4
5
9
7
7
9
2
2
5
3
1
8
4
2
8
3
6
4
7
9
5
1
9
2
4
1
3
5
6
7
8
1
5
7
8
2
9
3
6
4
meta-llama/llama-3.3-70b-instruct
meta
5
3
4
6
7
8
9
1
2
6
7
2
5
9
3
9
8
4
6
8
5
1
2
7
9
4
8
8
4
1
3
8
5
7
3
9
9
2
8
4
1
4
2
5
6
7
5
3
2
5
6
1
9
8
2
9
6
8
4
1
5
5
3
4
1
9
7
3
2
8
6
5
3
6
7
9
6
9
4
2
1
deepseek/deepseek-chat-v3-0324
deepseek
1
3
5
6
9
8
2
4
7
8
7
2
5
4
1
9
6
3
6
9
4
3
2
7
5
1
8
5
4
1
9
8
6
7
3
2
3
2
8
7
1
4
5
9
6
7
6
9
2
5
3
1
8
4
2
8
6
1
4
5
3
7
9
9
1
4
8
3
2
6
5
1
4
5
7
6
9
1
8
2
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.