TECHNOLOGY PROFILE · TECH-0296
whisper.cpp
C/C++ inference implementation for Whisper models.
Reference link only — not reviewed this edition. This profile preserves the prepared catalog’s original descriptions. Advantages depend on the workload; comparisons are editorial, not benchmark results.
Where it has an edge
Portable native execution supports local transcription without a Python training stack.
Best-fit problems
On-device speech recognition on supported CPU/GPU hardware.
Tradeoffs & cautions
Feature/model support differs from the reference; quantization can affect accuracy.
What to test before choosing
Test real microphones/noise/languages; measure word errors, diarization, turn-taking, latency and voice rights.
Runtime & hardware
Local CPU/GPU or realtime infrastructure
Access & licensing
Code/model/voice licenses vary
Execution
Local / self-hosted
Lifecycle
Not independently audited
Comparison basis
Editorial problem-fit judgement; no head-to-head benchmark performed
Connected concepts
Alternatives to investigate
faster-whisper · Whisper