@Mrwhosetheboss If you watch till the end of the video, they have clearly mentioned that the performance is measured against M1 Pro/Max. So how is misleading??
Didn’t expect an engagement bait from you @Mrwhosetheboss
Hard agree! been using Sonnet 4.5, and it was amazing! It was proactive and had a personality to it.
Switched to KLM 2.5 & OpenAI 5.1, to try things out & it was buggy. not at all a good experience.
Use Sonnet or Opus. Gives great results. only issue is I’m hitting limits 🥲
Real talk. If you think Clawdbot is still janky and buggy it's probably mostly because you're using a model that's too dumb.
E.g. GPT-5 keeps fumbling around. Change to Opus 4.5 and it gets DRAMATICALLY better across the board, both remembering things and working out how to create/use tools.