Released: 2026-08-01
openai/gpt-4o
openai
8
3
6
2
9
7
1
4
5
7
5
9
8
1
6
2
8
3
1
4
2
9
7
5
3
6
8
6
8
4
3
5
1
7
7
9
9
2
7
4
3
8
8
1
6
3
1
5
7
8
2
6
2
7
2
6
8
6
4
9
5
5
1
4
7
1
5
2
3
9
9
1
5
9
3
1
6
4
4
3
2
google/gemini-3-flash-preview
8
1
9
3
6
7
2
4
5
5
7
6
8
1
2
9
4
3
3
4
2
9
4
5
7
6
8
6
3
8
4
5
1
2
7
9
9
5
7
2
3
6
8
1
4
4
2
1
7
8
9
5
3
6
1
8
4
6
9
7
3
5
2
7
6
5
5
2
3
4
8
1
2
9
3
1
7
4
6
5
8
meta-llama/llama-3.3-70b-instruct
meta
8
6
5
4
9
7
1
4
2
3
4
1
2
1
5
9
8
6
2
8
9
1
6
3
5
7
4
6
5
4
8
5
2
7
3
9
1
3
7
9
3
4
8
5
6
9
2
6
5
8
1
4
1
3
5
1
3
6
4
9
2
9
8
4
9
8
3
2
6
1
6
5
7
9
2
1
5
8
3
4
1
mistralai/mistral-medium-3
mistral
8
1
6
3
2
7
9
4
5
9
4
5
8
1
6
2
7
3
3
7
2
9
4
5
1
6
8
6
3
4
2
5
1
3
7
9
1
5
7
4
3
9
8
2
6
2
8
9
7
8
6
4
3
5
4
2
8
6
7
3
5
9
1
7
6
1
5
9
3
2
8
4
5
9
3
1
6
4
7
2
8
deepseek/deepseek-chat-v3-0324
deepseek
8
3
9
2
6
7
1
4
5
4
7
5
8
1
9
6
2
3
1
6
2
9
4
5
7
3
8
6
8
1
4
5
2
3
7
9
9
2
7
6
3
1
8
5
4
3
5
4
7
8
9
2
1
6
7
1
8
6
2
4
9
5
3
2
4
6
5
9
3
4
8
1
5
9
3
1
7
8
4
6
2
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.