AssemblyAI STT 3.7 Realtime Agent vs Deepgram Flux English

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
  • Flux English3.90% · 108ms

AssemblyAI STT 3.7 Realtime Agent is ahead of Deepgram Flux English on both datasets together, the Pipecat Dataset, time to first text and 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 AgentFlux English
  • Overall WER2.41%4.00%
  • Pipecat Dataset WER1.68%3.90%
  • Ocular Dataset WER3.81%4.18%
  • Time to first text493ms791ms
  • Final-text delay47ms108ms

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 AgentFlux English
  • ELO rating10241001–1039 · rank 1–16956938–991 · rank 12–22
  • Expected win rate60%40%
  • Ocular turns WER · 148 draws of 19536 won11 won
  • Pipecat WER · 757 draws of 999209 won33 won
  • Semantic WER · 179 draws of 19511 won5 won
  • Phone line · 148 draws of 19738 won11 won
  • Call-center noise · 89 draws of 19490 won15 won
  • Babble noise · 45 draws of 196142 won9 won
  • TV noise · 102 draws of 19672 won22 won
  • Reverb · 133 draws of 19653 won10 won
  • Short replies · 47 draws of 6817 won4 won
  • Entities · 43 draws of 494 won2 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 Flux English 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 Deepgram Flux English 4.00%, so AssemblyAI STT 3.7 Realtime Agent makes fewer errors by 1.58 points. On the 300 licensed Ocular turns, AssemblyAI STT 3.7 Realtime Agent scores 1.68% and Deepgram Flux English 3.90%, so AssemblyAI STT 3.7 Realtime Agent makes fewer errors by 2.23 points. Word error rate counts incorrect, missing, and extra words, so lower is better.

  • Yes. AssemblyAI STT 3.7 Realtime Agent finishes the transcript 298ms sooner. AssemblyAI STT 3.7 Realtime Agent has the final transcript 493ms after the speaker stops and Deepgram Flux English 791ms, 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 298ms 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 Deepgram Flux English after 4.18%, at the median.

Test your voice agent the same way.

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