How fast tokens stream in after the first one — the throughput a user feels once generation has started.
Continue to AI University →