Glossary
Small language model
A compact language model that runs fast and cheap and can handle focused tasks, sometimes on your own hardware.
A small language model (SLM) is a language model with far fewer parameters than frontier models, built to run quickly and cheaply on modest hardware, including laptops, phones or a single server. SLMs do well on focused, well-defined tasks such as classification, extraction, routing and short drafting. They are weaker at broad, open-ended reasoning.
There is no fixed size line, and what counts as small keeps moving as models improve. What matters is the trade: lower cost, lower latency and easier private deployment, in exchange for less general ability.
SLMs often work best as one part of a system. A small model handles the routine steps, and a larger one is called only when needed, which is a pattern [model routing](/glossary/model-routing) supports. Fine-tuning a small model on your own task is also common.
Start in 60 seconds
Put this to work on your project.
Get an AI-assisted, architect-reviewed scope, cost range, and timeline. Free, no login.