Tokens per second
Also known as Tokens per second, AI Tokens per second
Definition
Tokens per second is the serving concept concerned with tokens per second during AI inference.
What tokens per second means
Tokens per second is the serving concept concerned with tokens per second during AI inference. In a serving system, define the measurement boundary before collecting data. State whether the worker is warm, which model and precision are active, how batching is configured, and whether the request is streamed. These details determine whether two similar measurements describe the same operating condition.
How it appears in a workload
A team can compare a quiet run with a burst of representative requests and observe how tokens per second changes. Record relevant timestamps and runtime signals with the request rather than relying on one dashboard aggregate. A useful comparison includes request mix, arrival pattern, concurrency, and output limits. For routed traffic, retain the selected route so a change can be investigated rather than attributed to tokens per second without evidence.
Limitations and tradeoffs
It should be interpreted with error rate and user-visible latency because an isolated gain can move cost or tail behavior elsewhere. Use it with p50 and tail latency, useful token throughput, errors, memory pressure, and cost. The best configuration is the one that meets the application's reliability and responsiveness requirements under representative traffic.
How this relates to Weave
Weave Router comparisons can record tokens per second beside model, provider, workload, and request outcome. That context helps explain whether a routing change altered tokens per second, but Weave does not make an inference runtime faster by itself. Keep serving conditions visible when using these observations to compare complete AI tasks.
Explore Router