Gradium vs GPT Realtime Whisper
Accuracy and speed
Left gets fewer words wrong. Lower has the final text ready sooner.
- Gradium6.48% · 212ms
- Realtime Whisper2.16% · 543ms
GPT Realtime Whisper is ahead on both datasets together, the Pipecat Dataset and time to first text, and Gradium on final-text delay. The Ocular Dataset is too close to call.
Last updated Methodology by Dileep Chagam
Who is better at what
- Overall WER5.78%2.63%
- Pipecat Dataset WER6.48%2.16%
- Ocular Dataset WER4.45%3.53%
- Time to first text1.71s1.38s
- Final-text delay212ms543ms
Lower is better on every measure. Too close to call means the likely ranges overlap. Our guide explains how word error rate is measured.
Gradium vs Realtime Whisper FAQ
Which is more accurate, Gradium or GPT Realtime Whisper?
GPT Realtime Whisper is more accurate on both datasets. On the 1,000 public Pipecat clips, Gradium scores 5.78% and GPT Realtime Whisper 2.63%, so GPT Realtime Whisper makes fewer errors by 3.15 points. On the eight licensed Ocular recordings, Gradium scores 6.48% and GPT Realtime Whisper 2.16%, so GPT Realtime Whisper makes fewer errors by 4.32 points. Word error rate counts incorrect, missing, and extra words, so lower is better.
GPT Realtime Whisper is more accurate on both datasets. On the 1,000 public Pipecat clips, Gradium scores 5.78% and GPT Realtime Whisper 2.63%, so GPT Realtime Whisper makes fewer errors by 3.15 points. On the eight licensed Ocular recordings, Gradium scores 6.48% and GPT Realtime Whisper 2.16%, so GPT Realtime Whisper makes fewer errors by 4.32 points. Word error rate counts incorrect, missing, and extra words, so lower is better.
No. GPT Realtime Whisper finishes the transcript 326ms sooner. Gradium has the final transcript 1.71s after the speaker stops and GPT Realtime Whisper 1.38s, at the median. Both are measured on the same 206 streamed turns.
GPT Realtime Whisper is the stronger pick for voice agents: it makes fewer errors on real conversations and finishes the transcript 326ms sooner. 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. Gradium shows first text after 4.45% and GPT Realtime Whisper after 3.53%, at the median.
Test your voice agent the same way.
Cekura runs these simulated calls on your agent and scores every one.