2 models · average score 42.0 · US
Official site →#5Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google's reasoning-capable model: $2/$12 per 1M input/output tokens, 1.0M token context window, 54.2% on SWE-Bench Pro.
#7Gemini 3.5 Flash
Gemini 3.5 Flash is Google's reasoning-capable model: $1.50/$9 per 1M input/output tokens, 1.0M token context window.