Released: 2026-06-25
openai/gpt-4o
openai
7
9
6
5
4
1
8
2
3
3
2
4
6
9
8
1
5
7
5
1
8
7
2
3
9
6
4
2
6
3
9
5
7
7
4
1
9
8
5
2
6
4
3
1
6
4
7
1
3
8
6
5
2
9
6
5
7
1
3
9
2
7
8
1
4
2
8
7
6
6
3
5
8
3
9
4
1
2
7
9
7
google/gemini-3-flash-preview
5
6
7
1
4
3
2
8
9
2
9
4
6
8
7
1
5
3
3
1
8
5
2
9
6
7
4
8
2
6
9
1
5
3
4
7
9
3
5
2
7
8
4
1
6
4
7
1
3
6
4
5
2
8
6
5
3
7
9
1
8
4
2
1
8
2
4
5
6
7
3
9
4
4
9
8
3
2
5
6
7
mistralai/mistral-medium-3
mistral
9
3
7
8
4
5
2
6
1
2
4
8
6
3
9
1
5
7
5
1
8
7
2
1
9
6
3
7
6
3
9
5
8
4
2
1
4
9
5
2
7
3
8
1
6
8
2
1
3
6
4
5
7
9
6
5
4
1
9
7
3
8
2
1
7
2
4
8
6
9
3
5
3
8
9
5
1
2
6
4
7
deepseek/deepseek-chat-v3-0324
deepseek
5
7
6
8
4
1
2
9
3
9
2
3
6
7
5
1
8
4
4
1
8
9
2
3
6
7
5
2
3
7
5
6
8
9
4
1
8
4
5
2
9
7
3
1
6
6
9
1
3
5
4
5
2
8
6
5
4
1
8
9
7
3
2
1
8
2
7
3
6
4
5
9
3
6
9
4
1
2
8
5
7
meta-llama/llama-3.3-70b-instruct
meta
5
3
4
6
7
8
9
1
2
6
7
2
1
9
5
3
4
8
1
9
8
3
4
2
5
6
7
8
5
9
7
6
1
4
2
3
4
2
6
8
5
3
7
9
1
7
1
3
9
2
4
8
5
6
9
6
1
5
3
7
2
8
4
2
8
7
4
1
9
6
3
5
3
4
5
2
8
6
1
7
9
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.