OpenAI is ahead on time to first text, and Zoom on final-text delay. The Pipecat Dataset and The Ocular Dataset are too close to call.
Last updated Methodology by Dileep Chagam
Each row sets OpenAI's best model against Zoom's best on that measure.
It is too close to call: the gap is inside the margin of error. The best OpenAI result is GPT Realtime Whisper at 2.16%, and the best Zoom result is Zoom Scribe Pro at 1.96%, on the Pipecat Dataset.
It is too close to call: the gap is inside the margin of error. The best OpenAI result is GPT Realtime Whisper at 2.16%, and the best Zoom result is Zoom Scribe Pro at 1.96%, on the Pipecat Dataset.
Zoom is faster, 425ms sooner. The best OpenAI result is GPT Realtime Whisper at 543ms, and the best Zoom result is Zoom Scribe at 118ms, on final-text delay.
OpenAI: GPT-4o Mini Transcribe, GPT-4o Transcribe, GPT Realtime Whisper. Zoom: Zoom Scribe, Zoom Scribe Pro. Every model ran the same tests on Converse-STT.