Released: 2026-05-12
google/gemini-3-flash-preview
4
1
5
3
2
9
6
8
7
6
8
3
5
1
7
4
9
2
7
9
2
8
4
6
3
1
5
3
6
4
7
9
2
5
1
8
2
5
1
4
6
8
9
7
3
9
7
8
1
4
5
2
3
6
8
3
7
6
2
1
0
5
4
1
4
6
9
5
3
7
2
8
5
2
9
6
7
4
1
6
3
openai/gpt-4o
openai
5
1
4
3
2
9
7
6
8
6
9
3
5
1
7
4
8
2
7
8
2
8
9
6
1
3
5
8
7
6
2
3
4
5
1
9
2
3
5
1
6
8
9
7
4
9
4
1
7
5
5
6
2
3
3
2
7
6
8
1
2
5
6
1
4
8
9
4
3
6
5
7
4
5
9
7
7
2
8
9
1
meta-llama/llama-3.3-70b-instruct
meta
5
1
9
7
2
8
6
3
4
6
8
3
5
9
7
2
4
1
7
4
2
8
6
1
5
9
3
3
5
1
9
4
2
7
8
6
2
9
8
1
5
6
4
7
3
4
7
6
3
8
5
1
2
9
9
3
5
6
1
4
8
7
2
8
2
4
7
3
9
5
1
6
1
6
7
4
5
3
9
2
8
google/gemini-2.0-flash-001
9
1
5
4
2
8
6
3
7
6
8
3
5
1
7
4
9
2
7
2
4
8
3
6
1
5
9
4
9
6
7
8
2
5
1
3
2
5
8
3
6
1
9
7
4
3
7
1
9
4
5
2
6
8
8
3
7
6
9
4
2
5
1
5
4
2
1
6
3
7
8
9
1
6
9
2
7
5
3
4
8
mistralai/mistral-medium-3
mistral
5
1
4
9
2
8
3
6
7
6
9
3
5
1
7
4
8
2
7
2
8
8
3
6
1
9
5
4
6
7
3
8
2
5
1
9
2
8
1
4
5
9
9
7
3
3
5
9
1
4
7
2
6
8
1
3
2
6
9
4
7
5
8
8
4
6
2
7
3
9
1
5
9
7
5
8
6
1
3
2
4
deepseek/deepseek-chat-v3-0324
deepseek
5
1
8
4
2
9
6
3
7
6
9
3
5
1
7
4
8
2
7
2
4
8
3
6
1
9
5
4
7
6
9
8
2
5
1
3
2
5
1
3
6
8
9
7
4
9
8
7
1
4
5
2
6
8
8
3
2
6
9
1
7
5
4
1
4
5
2
8
3
6
7
9
3
6
9
7
5
4
8
2
1
x-ai/grok-3-mini-beta
x-ai
5
1
6
9
2
8
7
4
3
6
8
3
5
1
7
4
9
2
7
2
4
8
3
6
1
9
5
4
9
2
3
6
5
8
1
7
2
3
1
4
8
9
6
7
5
1
7
5
2
4
3
9
8
6
8
3
7
6
9
4
2
5
1
9
4
2
1
5
3
8
6
7
3
5
9
7
1
6
4
2
8
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.