How do different operating systems and microarchitectures affect the end user experience with agentic AI?
We looked through 2.91M actual Claude tool use responses to find out.
By our data, it appears that Linux is typically faster than Windows by a factor of about 3x in execution of shell commands. (1/2)🧵
agentic loop — The cycle an agent runs in: decide, call a tool, read the result, decide again — repeating until the goal is met or a stop condition fires.
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
Why it matters
If shell latency dominates your agentic loopThe cycle an agent runs in: decide, call a tool, read the result, decide again — repeating until the goal is met or a stop condition fires.Full definition →, the host OS is a 3x lever measured on real traffic rather than a synthetic benchmarkA standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.Full definition →.
articleHow will AI change operating systems? Part 2: WindowsGergely Orosz
postHow often does Claude use tools?SemiAnalysis
articleIs Bash All You Need? An Empirical Study of Tool Interfaces for Enterprise Digital Worker AgentsHazel Mak, Susheel Suresh, Sahil Bhatnagar, Barry Wang, Chhaya Methani, Alejandro Gutierrez Munoz