Glossary term
Glossary term
Agentic Systems
A family of relatively small Gemini models optimized for speed and low latency. Flash models are designed for a wide range of applications where quick responses and high throughput are crucial.
Created for this library
A SaaS team uses a Flash model in its product chat because its latency budget for each suggestion is under one second.
An LLM product team picks a Flash model for high-volume autocomplete features where per-call cost matters more than maximum quality.
A startup uses a Flash model behind the scenes for its summary feature and routes only edge cases to a larger model.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License