Blog
Recent headlines.
- Quantized Whisper-small vs large-v2: 10dB Draft, Offline Win Sep 2, 2026
- Whisper at 680,000 Hours: WER Trade-offs Below 15 dB SNR Sep 1, 2026
- Synthetic TTS Data for Low-Resource ASR: A Wolof Case Study Aug 30, 2026
- Whisper large-v3 Across 6 Accents: What the Leaderboard Shows Aug 29, 2026
- Claude's 5-Hour Window vs. Batch API: Long-Audio Token Cost Aug 27, 2026
- 2026 Benchmark: HLS AAC Adds 2.6% WER, 1.1% DER vs WAV Aug 26, 2026
- Whisper-large-v3 vs Scribie: Two Scoreboards, No Winner Aug 24, 2026
- Rev.com Pay: What 'Up to $1/Minute' Actually Nets Per Hour Aug 23, 2026
- MIT Audit: Diarization Costs Valid; Three-Zone ROI Exposes Risks Aug 21, 2026
- MIT SLP 2026 Diarization: Embedding Drift & Pipeline Architecture Aug 20, 2026
- Conformer ASR: Front-Ends & LMs Drive Low-Resource Noise Gains Aug 18, 2026
- Diarization Cuts Podcast WER by 18.4%: Pre vs Post ASR Aug 16, 2026
- Whisper VAD: 32% Diarization Error Reduction Is Conditional Aug 15, 2026
- DER Drop to 10% Cuts Editing: Service B Wins by 43% Aug 13, 2026
- Rev.com Test-to-Wage: $36/Hour vs Minimum Wage Reality Aug 12, 2026
- DER's Fatal Flaw in Legal Depositions: Overlap-Aware Wins Aug 11, 2026