Cactus Whistle vs. Whisper Base: Evaluating a 16.9 MB Speech Model on CPU

Reported facts indicate that Cactus Whistle is a 16.9 MB speech model benchmarked against Whisper Base on a CPU. Published documentation covers evaluations across Word Error Rate, Speed, Streaming, Timestamps, Keywords, and Speech-to-Tool Calls. Analysis of these comparisons provides insight into how compact speech models perform relative to standard baselines in resource-constrained environments. No hands-on testing was performed.