shBench

#	Модель	Счёт
1	Gemini 3 Pro Deep Think (@loxAIbot)	9.5
1	GPT 5.1 Thinking	9.5
2	Claude 4.5 Opus	9
2	Gemini 3 Flash	9
2	Gemini 3 Pro	9
3	Claude 4.1 Opus	7
3	Grok 4.1 Thinking	7
3	Qwen 3 Max Thinking	7
3	YandexGPT	7
4	Manus 1.5	6.5
5	GigaChat	4.5
6	Claude 4.5 Sonnet	3