AssemblyAI STT 3.7 Realtime Agent 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)
  • STT 3.7 Realtime Agent1.68% · 47ms
  • STT-RT v52.54% · 38ms

AssemblyAI STT 3.7 Realtime Agent is ahead on both datasets together, the Pipecat Dataset and time to first text, and Soniox STT-RT v5 on final-text delay. The Ocular Dataset is too close to call.

Last updated Methodology by Dileep Chagam

Who is better at what

STT 3.7 Realtime AgentSTT-RT v5
  • Overall WER2.41%3.02%
  • Pipecat Dataset WER1.68%2.54%
  • Ocular Dataset WER3.81%3.92%
  • Time to first text493ms1.20s
  • Final-text delay47ms38ms

Lower is better on every measure. Too close to call means the likely ranges overlap.

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.

STT 3.7 Realtime AgentSTT-RT v5
  • ELO rating10241001–1039 · rank 1–161000986–1007 · rank 6–20
  • Expected win rate53%47%
  • Ocular turns WER · 155 draws of 19831 won12 won
  • Pipecat WER · 842 draws of 999107 won50 won
  • Semantic WER · 186 draws of 1988 won4 won
  • Phone line · 157 draws of 19831 won10 won
  • Call-center noise · 133 draws of 19743 won21 won
  • Babble noise · 110 draws of 19760 won27 won
  • TV noise · 133 draws of 19862 won3 won
  • Reverb · 143 draws of 19838 won17 won
  • Short replies · 52 draws of 7015 won3 won
  • Entities · 47 draws of 503 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

STT 3.7 Realtime Agent vs STT-RT v5 FAQ

  • AssemblyAI STT 3.7 Realtime Agent is more accurate on both datasets. On the 1,000 public Pipecat clips, AssemblyAI STT 3.7 Realtime Agent scores 2.41% and Soniox STT-RT v5 3.02%, so AssemblyAI STT 3.7 Realtime Agent makes fewer errors by 0.61 points. On the 300 licensed Ocular turns, AssemblyAI STT 3.7 Realtime Agent scores 1.68% and Soniox STT-RT v5 2.54%, so AssemblyAI STT 3.7 Realtime Agent makes fewer errors by 0.87 points. Word error rate counts incorrect, missing, and extra words, so lower is better.

  • Yes. AssemblyAI STT 3.7 Realtime Agent finishes the transcript 702ms sooner. AssemblyAI STT 3.7 Realtime Agent has the final transcript 493ms after the speaker stops and Soniox STT-RT v5 1.20s, at the median. Both are measured on the same 206 streamed turns.

  • AssemblyAI STT 3.7 Realtime Agent is the stronger pick for voice agents: it makes fewer errors on real conversations and finishes the transcript 702ms 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. AssemblyAI STT 3.7 Realtime Agent shows first text after 3.81% and Soniox STT-RT v5 after 3.92%, at the median.

Test your voice agent the same way.

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