๐ง Google Gemini 2.5 Pro Gets Deep Think Mode โ 82% on GPQA Diamond
Google's Gemini 2.5 Pro now ships with a parallel-thinking Deep Think reasoning mode, scoring 82.4% on GPQA Diamond and 89.8% on MMLU-Pro with a 2 million token context window. That puts it squarely in the frontier reasoning tier alongside the latest from OpenAI and Anthropic. Is this the reasoning model to beat right now?
โก 26 buzzing
313 views
๐ 0 subscribed