AssemblyAI Universal 3.5 Pro 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)
  • Universal 3.5 Pro1.93% · 180ms
  • STT-RT v52.54% · 38ms

AssemblyAI Universal 3.5 Pro 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. AssemblyAI Universal 3.5 Pro has a newer release: AssemblyAI Universal 3.6 Pro vs Soniox STT-RT v5.

Last updated Methodology by Dileep Chagam

Who is better at what

Universal 3.5 ProSTT-RT v5
  • Overall WER2.72%3.02%
  • Pipecat Dataset WER1.93%2.54%
  • Ocular Dataset WER4.23%3.92%
  • Time to first text489ms1.20s
  • Final-text delay180ms38ms

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.

Universal 3.5 ProSTT-RT v5
  • ELO rating1017993–1033 · rank 2–181000986–1007 · rank 6–20
  • Expected win rate53%47%
  • Ocular turns WER · 154 draws of 19728 won15 won
  • Pipecat WER · 833 draws of 999102 won64 won
  • Semantic WER · 183 draws of 1979 won5 won
  • Phone line · 148 draws of 19831 won19 won
  • Call-center noise · 135 draws of 19837 won26 won
  • Babble noise · 108 draws of 19859 won31 won
  • TV noise · 138 draws of 19851 won9 won
  • Reverb · 136 draws of 19834 won28 won
  • Short replies · 51 draws of 6915 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

Universal 3.5 Pro vs STT-RT v5 FAQ

  • AssemblyAI Universal 3.5 Pro is more accurate on both datasets. On the 1,000 public Pipecat clips, AssemblyAI Universal 3.5 Pro scores 2.72% and Soniox STT-RT v5 3.02%, so AssemblyAI Universal 3.5 Pro makes fewer errors by 0.30 points. On the 300 licensed Ocular turns, AssemblyAI Universal 3.5 Pro scores 1.93% and Soniox STT-RT v5 2.54%, so AssemblyAI Universal 3.5 Pro makes fewer errors by 0.62 points. Word error rate counts incorrect, missing, and extra words, so lower is better.

  • Yes. AssemblyAI Universal 3.5 Pro finishes the transcript 706ms sooner. AssemblyAI Universal 3.5 Pro has the final transcript 489ms after the speaker stops and Soniox STT-RT v5 1.20s, at the median. Both are measured on the same 206 streamed turns.

  • AssemblyAI Universal 3.5 Pro is the stronger pick for voice agents: it makes fewer errors on real conversations and finishes the transcript 706ms 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 Universal 3.5 Pro shows first text after 4.23% 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.