Released: 2026-06-08
google/gemini-3-flash-preview
1
5
3
7
4
2
8
6
9
4
8
7
1
6
9
2
5
3
2
6
9
3
8
5
7
1
4
7
4
5
2
9
1
3
8
6
8
2
0
6
3
4
1
9
7
3
9
6
5
7
8
4
2
1
6
3
1
9
5
7
0
4
8
9
1
2
8
0
3
6
7
5
5
7
8
4
2
6
9
3
1
openai/gpt-4o
openai
5
1
3
7
4
2
8
6
9
4
8
7
1
6
9
5
2
3
2
6
9
3
8
5
7
1
4
7
4
5
2
9
8
3
5
6
8
9
6
6
3
4
1
8
2
3
2
1
5
7
8
4
9
7
9
5
8
6
5
7
2
4
1
6
3
2
9
1
3
6
7
5
1
7
4
4
2
6
9
3
8
mistralai/mistral-medium-3
mistral
4
5
3
2
6
7
8
1
9
7
8
6
1
3
9
4
2
5
2
6
9
5
8
4
3
7
1
3
4
5
7
9
1
2
8
6
8
9
7
6
2
4
1
5
3
1
2
6
8
7
3
9
4
5
6
7
8
9
5
2
4
3
1
5
1
2
4
3
3
6
7
8
9
3
4
6
1
8
5
2
7
meta-llama/llama-3.3-70b-instruct
meta
5
1
3
7
8
9
4
2
6
4
8
9
1
6
5
3
7
2
2
6
7
3
4
8
5
9
1
9
5
1
8
2
6
7
3
4
8
3
4
5
1
7
2
6
9
7
9
6
4
3
2
1
8
5
1
4
5
9
7
3
6
8
2
3
2
8
6
5
1
9
4
7
6
7
2
4
9
8
5
1
3
deepseek/deepseek-chat-v3-0324
deepseek
1
7
3
2
6
8
8
5
9
4
8
5
1
3
9
2
6
7
2
6
9
5
4
7
3
1
8
3
2
5
8
9
1
7
4
6
8
9
7
6
2
4
1
3
5
6
4
1
3
7
5
8
2
9
9
3
8
7
5
2
6
4
1
6
1
2
9
8
3
4
7
5
7
5
4
4
1
6
9
8
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.