GPT-4o Mini Transcribe vs Soniox STT-RT v5

Compare

Accuracy and speed

Left gets fewer words wrong. Lower has the final text ready sooner.

Accurate and fast0ms200ms400ms600ms800ms0%2%4%6%Words wrong on clean audio (Pipecat WER)
  • 4o Mini Transcribe3.35% · 630ms
  • STT-RT v52.54% · 38ms

Soniox STT-RT v5 is ahead of GPT-4o Mini Transcribe on both datasets together, the Pipecat Dataset and final-text delay. The Ocular Dataset is too close to call.

Last updated Methodology by Dileep Chagam

Who is better at what

4o Mini TranscribeSTT-RT v5
  • Overall WER3.73%3.02%
  • Pipecat Dataset WER3.35%2.54%
  • Ocular Dataset WER4.45%3.92%
  • Time to first textn/a1.20s
  • Final-text delay630ms38ms

Lower is better on every measure. Too close to call means the likely ranges overlap. Our guide covers testing OpenAI Realtime API voice agents.

Clip by clip

On every clip both models were scored on, the lower WER wins, and WERs within 5 points of each other are a draw; in the rating, a wider win counts for more. Expected win rate is what the two ELO ratings predict for a match.

4o Mini TranscribeSTT-RT v5
  • ELO rating1001993–1011 · rank 6–181000986–1007 · rank 6–20
  • Expected win rate50%50%
  • Ocular turns WER · 151 draws of 19727 won19 won
  • Pipecat WER · 826 draws of 99995 won78 won
  • Semantic WER · 182 draws of 1979 won6 won
  • Phone line · 149 draws of 19828 won21 won
  • Call-center noise · 113 draws of 19837 won48 won
  • Babble noise · 98 draws of 19849 won51 won
  • TV noise · 79 draws of 195107 won9 won
  • Reverb · 126 draws of 19831 won41 won
  • Short replies · 46 draws of 7015 won9 won
  • Entities · 46 draws of 493 won0 won

Short replies are decided on the error count and entities on exact capture, with no draw margin. How the rating is worked out

4o Mini Transcribe vs STT-RT v5 FAQ

  • Soniox STT-RT v5 is more accurate on both datasets. On the 1,000 public Pipecat clips, GPT-4o Mini Transcribe scores 3.73% and Soniox STT-RT v5 3.02%, so Soniox STT-RT v5 makes fewer errors by 0.71 points. On the 300 licensed Ocular turns, GPT-4o Mini Transcribe scores 3.35% and Soniox STT-RT v5 2.54%, so Soniox STT-RT v5 makes fewer errors by 0.81 points. Word error rate counts incorrect, missing, and extra words, so lower is better.

  • Final-text delay was not measured for one or both, so speed cannot be compared. Both are measured on the same 206 streamed turns.

  • Soniox STT-RT v5 makes fewer errors on real conversations, and the two are not separated on speed. A voice agent cannot reply until the final transcript arrives, and the Ocular recordings are real conversations.

  • Only Soniox STT-RT v5. GPT-4o Mini Transcribe returns no text until the speaker stops, which matters if your agent reacts to partial transcripts.

Test your voice agent the same way.

Cekura runs these simulated calls on your agent and scores every one.