Glossary
Throughput
The volume of tasks or requests an AI agent can successfully process per unit of time, measuring operational capacity.
What is Throughput?
Throughput quantifies agent productivity and scalability, typically measured in tasks per second, requests per minute, or transactions per hour. High throughput enables agents to handle large workloads efficiently, while low throughput creates bottlenecks. Throughput depends on factors like model efficiency, infrastructure resources, task complexity, and implementation optimization.
Throughput must be considered alongside quality metrics since maximizing speed while sacrificing accuracy provides little value. Effective agents optimize the throughput-quality tradeoff, delivering maximum sustainable volume while maintaining acceptable quality standards. Throughput also interacts with cost, as increasing capacity often requires additional infrastructure investment.
Example
A document classification agent processes 500 documents per minute with 95% accuracy. After optimization, throughput increases to 800 documents per minute while maintaining 94% accuracy, enabling the same infrastructure to handle 60% more workload with minimal quality tradeoff.
How Signet addresses this
Signet considers throughput in the Reliability dimension as evidence of scalability and performance consistency. Agents demonstrating high, stable throughput under varying loads show operational maturity and earn stronger Reliability scores than those with lower or more variable capacity.
Build trust into your agents
Register your agents with Signet to receive a permanent identity and trust score.