Grok-3-Thinking based on the Gemini 2.5 Pro model with very similar reasoning and results. Try our Kimi 1.5K model in my profile