Released: 2026-06-02
openai/gpt-4o
openai
9
3
8
6
7
4
1
2
5
5
7
1
2
3
9
6
4
8
4
2
6
8
5
1
8
7
3
3
6
2
9
8
7
5
6
1
7
8
4
1
4
5
3
9
2
1
5
9
3
6
2
7
8
4
6
4
5
7
1
3
9
8
2
8
1
3
5
9
6
2
3
7
2
9
7
4
2
8
4
1
6
google/gemini-3-flash-preview
1
3
8
4
9
7
6
2
5
5
7
2
6
3
8
1
4
9
9
4
6
2
5
1
8
7
3
7
6
4
9
8
2
3
5
1
3
2
9
1
4
5
7
6
8
1
8
5
3
6
9
7
2
4
4
5
1
7
6
3
9
8
2
6
1
3
8
2
4
5
9
7
8
9
7
5
1
6
4
3
2
meta-llama/llama-3.3-70b-instruct
meta
5
7
8
3
9
6
1
4
2
4
1
3
8
2
5
9
6
7
9
6
2
4
5
1
8
3
7
8
5
9
7
6
4
2
1
3
3
4
1
2
8
9
5
7
6
2
8
6
5
1
7
4
9
3
1
9
5
6
3
8
7
2
4
7
3
4
9
1
2
6
5
8
6
2
7
1
4
3
9
8
5
mistralai/mistral-medium-3
mistral
4
3
8
6
7
9
5
2
1
5
7
2
3
1
8
6
4
9
9
6
1
4
5
2
8
7
3
8
6
3
9
2
7
5
1
4
2
5
9
1
4
3
8
6
7
7
1
4
8
6
5
2
9
3
1
4
5
7
9
3
6
8
2
6
2
3
5
8
1
9
4
7
3
9
7
2
4
6
1
5
8
deepseek/deepseek-chat-v3-0324
deepseek
4
3
8
6
9
7
1
2
5
5
7
2
3
1
4
9
6
8
6
9
1
5
2
8
4
7
3
2
6
5
9
3
8
7
4
1
7
8
3
1
4
5
2
9
6
9
1
4
2
6
7
5
3
8
1
4
6
7
8
3
9
5
2
3
2
9
4
5
6
8
1
7
8
5
7
9
2
1
6
3
4
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.