Speech recognition
6 RohitAI articles on Speech recognition: news, explainers and practical guides, newest first.
New Egyptian and Nigerian English tests broaden transcription comparisons
Hugging Face adds Egyptian and Nigerian English to its speech benchmark. What the grouped scores tell transcription buyers—and what remains unknown.
Choosing local transcription for Apple apps with AnySpeechRecognizer
Compare AnySpeechRecognizer’s local models for live captions, languages and speaker labels, with download sizes and iPhone-testing limits.
NVIDIA releases an open transcription model for recordings in 40 languages
Nemotron 3.5 Transcribe offers multilingual recording transcription, with different limits for translation, speaker labels and hosted access.
Microsoft adds 11 Indic languages to Azure LLM Speech transcription
Azure LLM Speech adds Tamil, Bengali and nine other Indic languages for transcription. Translation outputs, regional access and billing need separate checks.
IBM bundles transcription, chat and request sorting in Granite Switch
IBM’s new Granite Switch bundles combine speech transcription, chat and classification. Here is what they offer teams building voice-message intake.
Falcon-ASR for Arabic and Emirati Speech: What to Evaluate Now
Evaluate Falcon-ASR’s Arabic and Emirati speech claims, short-audio demo limits and benchmark evidence while API access remains planned.