OpenClaw vs Hermes - which is better?

I get that question a lot, but the real question is: better at what?
Well, how about poker?
Yep, I had my OpenClaw agent play my Hermes agent in a battle of Texas Hold'em
I made the setup interesting - they used inference credits from Openrouter as currency
But each decision used inference, so they also had to be smart and not waste their stack thinking too hard
As each won a hand, the winner's credit limit was increased while the loser's was decreased
So in theory, one agent could double their inference budget, minus whatever was used for inference in playing the game
So who won?
Hermes!
A few interesting stats:
>game was set for max of 100 hands
>$5 buy-in
>$0.05 / $0.10 blinds
>Hermes busted OpenClaw in hand 23
>165 total actions between both models
>avg decision time 3.36 secs
>longest decision 18 secs
Of course this was just one match, I plan to do more and will mix up the models and conditions and try to build a bigger dataset
This was just an interesting way to pit both agents against each other to see how they performed out of the box (neither agent received any training or skills related to poker prior to the match)