AssemblyAI STT 3.7 Realtime Agent vs Reson8 Rumble-2

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
  • Rumble-21.69% · 18ms

Reson8 Rumble-2 is ahead of AssemblyAI STT 3.7 Realtime Agent on both datasets together, time to first text and final-text delay. 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 AgentRumble-2
  • Overall WER2.41%2.05%
  • Pipecat Dataset WER1.68%1.69%
  • Ocular Dataset WER3.81%2.75%
  • Time to first text493ms486ms
  • Final-text delay47ms18ms

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 AgentRumble-2
  • ELO rating10241001–1039 · rank 1–161024990–1044 · rank 1–19
  • Expected win rate50%50%
  • Ocular turns WER · 168 draws of 19716 won13 won
  • Pipecat WER · 913 draws of 99939 won47 won
  • Semantic WER · 187 draws of 1977 won3 won
  • Phone line · 167 draws of 19818 won13 won
  • Call-center noise · 124 draws of 19746 won27 won
  • Babble noise · 124 draws of 19747 won26 won
  • TV noise · 155 draws of 19814 won29 won
  • Reverb · 164 draws of 19819 won15 won
  • Short replies · 60 draws of 708 won2 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 Rumble-2 FAQ

  • Reson8 Rumble-2 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 Rumble-2 2.05%, so Reson8 Rumble-2 makes fewer errors by 0.36 points. On the 300 licensed Ocular turns, AssemblyAI STT 3.7 Realtime Agent scores 1.68% and Reson8 Rumble-2 1.69%, with likely ranges of 1.47 to 1.90% and 1.48 to 1.92%. The ranges overlap, so the difference is too close to call. Word error rate counts incorrect, missing, and extra words, so lower is better.

  • No. Reson8 Rumble-2 finishes the transcript 7ms sooner. AssemblyAI STT 3.7 Realtime Agent has the final transcript 493ms after the speaker stops and Reson8 Rumble-2 486ms, at the median. Both are measured on the same 206 streamed turns.

  • Reson8 Rumble-2 finishes the transcript 7ms 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 Rumble-2 after 2.75%, at the median.

Test your voice agent the same way.

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