Running AI models on local devices rather than cloud servers. Edge deployment affects data residency, latency, security, and update mechanisms; may be preferred for privacy-sensitive applications.
See: Data residency; Latency; On-prem deployment; Small Language Model