GPT Realtime WhispervsZoom Scribe
Accuracy and speed
Left gets fewer words wrong. Lower has the final text ready sooner.
- Realtime Whisper2.16% · 543ms
- Scribe1.97% · 118ms
GPT Realtime Whisper is ahead on time to first text, and Zoom Scribe on final-text delay. The Pipecat Dataset and The Ocular Dataset are too close to call.
Last updated Methodology by Dileep Chagam
Who is better at what
- Pipecat Dataset WER2.16%1.97%
- Ocular Dataset WER3.53%4.88%
- Time to first text1.38s1.67s
- Final-text delay543ms118ms
Lower is better on every measure. Too close to call means the likely ranges overlap.
Related comparisons
Both providers, every model: OpenAI vs Zoom
Realtime Whisper vs Scribe FAQ
Which is more accurate, GPT Realtime Whisper or Zoom Scribe?
GPT Realtime Whisper and Zoom Scribe are too close to call on accuracy. On the 1,000 public Pipecat clips, GPT Realtime Whisper scores 2.16% and Zoom Scribe 1.97%, with likely ranges of 1.92 to 2.41% and 1.73 to 2.22%. The ranges overlap, so the difference is too close to call. On the eight licensed Ocular recordings, GPT Realtime Whisper scores 3.53% and Zoom Scribe 4.88%, with likely ranges of 2.58 to 4.51% and 4.41 to 5.29%. The ranges overlap, so the difference is too close to call. Word error rate counts incorrect, missing, and extra words, so lower is better.
GPT Realtime Whisper and Zoom Scribe are too close to call on accuracy. On the 1,000 public Pipecat clips, GPT Realtime Whisper scores 2.16% and Zoom Scribe 1.97%, with likely ranges of 1.92 to 2.41% and 1.73 to 2.22%. The ranges overlap, so the difference is too close to call. On the eight licensed Ocular recordings, GPT Realtime Whisper scores 3.53% and Zoom Scribe 4.88%, with likely ranges of 2.58 to 4.51% and 4.41 to 5.29%. The ranges overlap, so the difference is too close to call. Word error rate counts incorrect, missing, and extra words, so lower is better.
No. Zoom Scribe finishes the transcript 425ms sooner. GPT Realtime Whisper has the final transcript 543ms after the speaker stops and Zoom Scribe 118ms, at the median. Both are measured on the same 206 streamed turns.
Zoom Scribe finishes the transcript 425ms sooner, and the two are too close to call on real conversations. A voice agent cannot reply until the final transcript arrives, and the Ocular recordings are real conversations.
Yes, both return partial text before the speaker stops. GPT Realtime Whisper shows first text after 1.38s and Zoom Scribe after 1.67s, at the median, so GPT Realtime Whisper is 289ms sooner.