AssemblyAI STT 3.7 Realtime Agent vs Reson8 Resonant-1

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
  • Resonant-12.10% · 235ms

AssemblyAI STT 3.7 Realtime Agent is ahead on time to first text and final-text delay, and Reson8 Resonant-1 on both datasets together. The Pipecat Dataset and The Ocular Dataset are too close to call.

Last updated Methodology by Dileep Chagam

Who is better at what

STT 3.7 Realtime AgentResonant-1
  • Overall WER2.41%2.38%
  • Pipecat Dataset WER1.68%2.10%
  • Ocular Dataset WER3.81%2.93%
  • Time to first text493ms2.40s
  • Final-text delay47ms235ms

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 AgentResonant-1
  • ELO rating10241001–1039 · rank 1–1610341021–1047 · rank 1–10
  • Expected win rate49%51%
  • Ocular turns WER · 177 draws of 1987 won14 won
  • Pipecat WER · 890 draws of 99967 won42 won
  • Semantic WER · 191 draws of 1982 won5 won
  • Phone line · 175 draws of 1989 won14 won
  • Call-center noise · 127 draws of 19739 won31 won
  • Babble noise · 138 draws of 19739 won20 won
  • TV noise · 154 draws of 19723 won20 won
  • Reverb · 163 draws of 19815 won20 won
  • Short replies · 64 draws of 702 won4 won
  • Entities · 48 draws of 502 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 Resonant-1 FAQ

  • Reson8 Resonant-1 is more accurate on the public clips. On the 1,000 public Pipecat clips, AssemblyAI STT 3.7 Realtime Agent scores 2.41% and Reson8 Resonant-1 2.38%, so Reson8 Resonant-1 makes fewer errors by 0.03 points. On the 300 licensed Ocular turns, AssemblyAI STT 3.7 Realtime Agent scores 1.68% and Reson8 Resonant-1 2.10%, with likely ranges of 1.47 to 1.90% and 1.86 to 2.35%. The ranges overlap, so the difference is too close to call. Word error rate counts incorrect, missing, and extra words, so lower is better.

  • Yes. AssemblyAI STT 3.7 Realtime Agent finishes the transcript 1.91s sooner. AssemblyAI STT 3.7 Realtime Agent has the final transcript 493ms after the speaker stops and Reson8 Resonant-1 2.40s, at the median. Both are measured on the same 206 streamed turns.

  • AssemblyAI STT 3.7 Realtime Agent finishes the transcript 1.91s sooner, and the two are too close to call on real conversations. 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 Reson8 Resonant-1 after 2.93%, at the median.

Test your voice agent the same way.

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