Released: 2026-05-18
openai/gpt-4o
openai
4
5
2
7
9
1
6
3
8
6
3
4
5
2
8
1
9
7
1
9
7
3
6
4
5
4
2
5
4
3
9
1
6
7
8
2
7
6
9
4
7
5
8
1
3
8
2
1
2
5
3
9
7
4
2
8
6
6
3
7
4
9
1
3
7
5
1
8
2
2
6
5
9
1
8
6
4
9
3
2
6
google/gemini-2.0-flash-001
4
3
5
7
2
1
6
9
8
6
9
8
5
4
3
1
2
7
2
7
1
6
9
8
5
4
3
7
2
3
4
5
6
9
8
1
1
6
9
8
7
2
4
5
3
8
5
4
2
1
9
3
7
6
5
4
2
3
6
7
8
1
9
3
7
6
1
8
5
2
3
4
9
1
7
9
3
4
7
6
5
google/gemini-3-flash-preview
4
2
5
7
9
1
6
3
8
6
3
7
4
5
8
1
2
9
9
8
1
6
2
3
5
4
7
7
5
3
9
1
6
8
4
2
2
4
9
8
7
5
1
6
3
1
6
8
2
4
3
9
7
5
5
4
2
5
3
7
2
9
1
3
7
4
1
6
5
2
8
9
8
1
6
3
5
4
7
2
6
meta-llama/llama-3.3-70b-instruct
meta
4
5
7
7
3
1
6
2
8
6
8
2
9
5
4
1
3
5
2
3
1
8
6
5
5
4
9
8
2
3
5
1
6
4
9
1
5
9
9
3
7
8
2
1
3
1
6
4
2
9
3
8
7
5
3
1
5
4
2
9
7
8
6
9
7
8
1
4
5
3
6
2
7
4
6
6
8
2
9
5
1
mistralai/mistral-medium-3
mistral
4
2
5
7
9
1
6
3
8
6
8
7
3
4
2
1
5
9
1
9
3
8
6
5
5
4
2
7
4
3
9
5
6
8
2
1
5
6
9
4
7
8
2
1
3
8
1
2
2
3
9
4
7
6
2
5
4
6
3
7
9
9
1
9
7
6
1
8
5
3
4
4
3
1
8
5
2
4
7
6
5
deepseek/deepseek-chat-v3-0324
deepseek
4
3
1
7
5
9
6
2
8
6
2
7
4
8
3
1
5
9
9
5
8
6
2
1
5
4
3
1
8
3
5
4
6
9
7
2
2
4
9
8
7
1
5
6
3
5
6
4
2
9
8
3
7
1
7
9
2
3
6
5
8
1
4
8
7
5
1
3
4
2
9
6
3
1
6
9
8
2
4
3
5
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.