Transcript Features
Deepgram Speech AI Platform · Deepgram
Speech AI Platform — Deepgram (Nova STT, Aura TTS, Voice Agent)
Deepgram evals — Transcript Features (relift v3 InfraRed)
About Deepgram
Deepgram is a speech-AI platform offering streaming and batch speech-to-text (Nova), Aura text-to-speech, speaker diarization, redaction, and smart formatting across 30+ languages — used by voice-agent platforms, contact centers, and media teams.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Operator sets punctuate=true to get sentence-terminated, capitalized transcripts. | Set punctuate=true; the response transcript includes punctuation and capitalization in words[].punctuated_word (where present) and alternatives[].transcript. Consumers that need raw lowercase tokens must explicitly opt out (punctuate=false), not strip punctuation client-side which loses sentence bo… | Pass / FailAi Platformmedium |
| 02 | Operator wants digit form for numbers ('123') instead of spelled form ('one twenty-three'). | Set numerals=true (a lighter formatting flag than smart_format). On the same audio, numerals affects only number token form. If both smart_format=true and numerals=true are sent, smart_format wins for entities — verify on the specific audio domain before pinning both. | Pass / FailAi Platformlow |
| 03 | Operator transcribes a sales call mentioning phone numbers, dates, and currency. | Set smart_format=true to get formatted entities (phone numbers as (415) 555-1234, dates as 'June 15, 2026', currency as $1,234.56). smart_format implies punctuate=true. Verify formatted output against your downstream parsing — formatted entities are strings, not structured fields. | Pass / FailAi Platformmedium |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Deepgram
- Ai Platform
- Transcript Features
Recommended for
Works with
Related evals
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.