let them cook → choose your chef

Bro it doesn't matter how many chefs are in the kitchen... If the price of the meal gonna be most more than buying a private chef.

depends on the meal

What meal are we serving today? Love to see some more examples! (Not trolling, genuinely curious)

So burn 4x as many tokens on the same task? Throw away three of them? I don’t understand. What is the advantage here?

Why not use the best chef for every task?

different strengths, same kitchen

xD

i only wish someone would give me an ultra plan 😮‍💨 :whistle: >:P

Oh this is really good. It could use a consolidation step which picks the best parts of each implementation.

This feature is so underrated man

i mean, haven't you heard of "too many cooks spoil the broth"

What's this ?!

They should work together

Almost cancelled my plan today, I guess not I’m here to stay

Fun idea

Whoa.

How can I create/edit demo videos like this?

Do you pay for all of them?

finally, a way to get four different hallucinations for the price of five.

poor worktree design for parallel feature development. copy Codex pls

love your work man!

I haven’t found this feature useful I find it hard to evaluate the end result without feeling like I waste more time then I gain in value I also burn more tokens…

I still don't find a use case for using multiple models for a task.

this probably costs a kidney or smth are you ok?

200 usd and we got nothing when compared to Claude code.

Bro wants us to bankrupt.

Where's the option to "cross-review approaches, cherry pick and distill onto optimal solutions and review with operator"?

Hey Erik! Do you need creators to scale your brand organically? Run a creator academy and would love to set you up with some of our best for free. Sent you a DM

The "Usage Shredder 3000™"

People still use Cursor?

this is very interesting

who u picking to cook for u? 😂

4 model to handle div centering is good use?

RIP BlackBox AI ig

What a token sink

The last time I tried the #x agents thing it borked my git repo completely. What exactly is the purpose of this other than burning tokens? It looked like it was using worktrees which means none of them had proper env or local resources to do any testing, so nothing actually workd

is cursor still a thing?

Will that work well on the 20$ plan?