the Grok @bot desktop app just got a whole lot faster! there's still more to improve, but we shipped many perf improvements over the past week:
• cold start to usable shell: -247ms (18% faster).
• cold start LCP: -425ms (23% faster).
• cold start FCP: -234ms (16% faster).
•
The @bot team has been been poring over the harness the past 2 weeks and we've found many opportunities to improve our routing, caching, dynamic context usage, and overall token efficiency.
Your usage should go much further now. On average 10%, and up to 35% more effective usage
send me all your Grok @bot bugs with the in-app Send Feedback! i'll get my botetos to investigate them, just mention "poteto poteto poteto" in the feedback
you can also ask your bot to send feedback
