The system prompt is injected into your context on the server side.
here is a collection of public system prompts from claude code: https://github.com/Piebald-AI/claude-code-system-prompts
...the absolute state of the token maximizers.
I'm assuming this is a sarcastic response to the person who said "640K [RAM] ought to be enough for anybody!"
Because, as somebody who was around when we had 640 K RAM, people certainly weren't happy with that amount.
Although as soon as I wrote 'foreseeable future' I came to the realization that this is far far far less far out than it used to be. Which might mean I simply agree with you.
Checked yesterday, for that day alone I had used $168 worth on my $20 subscription in Claude Code. I still had plenty of weekly use left. Seems like subscriptions are discounted at a 1:10 rate?
EDIT: That was a delightfully thorough analysis. Nice to see Anthropic taking the value crown, only because Opus 5.5 is such a joy to use.
Because a big part of the price difference is due to Anthropic's models being more expensive that OpenAIs AND using more tokens for the same tasks (at least according to artificialanalysis.ai).
I use a personal Claude account for personal projects and can let rabl run for an hour and barely make a dent into my usage
On the enterprise I have to be a lot more careful or I can burn through 2k in a week
The excuse they give is the guarantees you get with enterprise plans that they won’t look at your data