Apple's new SpeechAnalyzer API, benchmarked against Whisper and its predecessor
Independent testing shows that Apple's new SpeechAnalyzer API significantly outperforms its predecessor and various Whisper models in speech-to-text accuracy. The new engine is faster and more accurate on both clean and noisy audio, providing a clear upgrade for developers.
Why it matters
This benchmark provides critical data for developers choosing between proprietary on-device AI tools and open-source alternatives like Whisper.
Apple shipped SpeechAnalyzer with no accuracy numbers. We measured it against the API it replaces and against Whisper, on 5,559 standard test utterances, and released every transcript.
Apple's new SpeechAnalyzer is the most accurate on-device speech engine we tested. It beat every Whisper model we ship, including Whisper Small, on both the clean and the noisy half of LibriSpeech, while running roughly three times faster than Small. And the API it replaces, SFSpeechRecognizer, came last on clean speech: behind even Whisper Tiny, a 40MB model.
Lower is better: WER is word error rate, the percentage of words an engine substitutes, drops, or invents. LibriSpeech test-clean is 2,620 utterances of clean read speech; test-other is 2,939 harder, noisier utterances. Every engine ran fully on-device on an Apple M2 Pro (32GB, macOS 26.5.1).
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in