Improvement
Babbily has removed output token restrictions across the platform. Every response, from GPT-4.1 to Claude 4 Opus and Gemini 2.5, now uses the maximum output each model allows.
Longer, more detailed answers
Better context retention
Fewer replies cut off mid-thought