Edge AI
Also known as: edge inference
Edge AI is the umbrella term for inference that happens on the device or the local network, rather than in a remote data center.
It includes on-device LLMs, local RAG, and private deployments. The common thread: the data and the compute stay close.
- Lower latency, works offline.
- Privacy and data-sovereignty.
- Bounded by local hardware.
Why it's growing: cheaper hardware + smaller-capable models mean "good enough" AI increasingly doesn't need the cloud. Edge AI is that shift in a word.