HomeToolsCategoriesModel watchFuelGoldInsightsDailyHow we vetContact
Submit a tool

Deepgram

Speech-to-text built for noisy recordings and hour-long calls.

AI Audio & VoiceFreemium★4.5 editor score

What it does

Deepgram returns word timings, speaker labels and keyword boosting for jargon — part numbers, brand names, port names. On long files it is faster and cheaper than most general-purpose models.

The boost list is what makes it usable for trade work. Without it, 'FOB Shenzhen' arrives as three unrelated words and the transcript is worthless.

Standout features

  • Speaker diarisation
  • Keyword boosting
  • Batch and streaming
  • 35+ languages

Best for

TranscriptionSpeechAPI