AssemblyAI Universal 3.6 Pro vs NVIDIA Nemotron 3.5 Transcribe
Accuracy and speed
Left gets fewer words wrong. Lower has the final text ready sooner. Nemotron 3.5 Transcribe doesn't support streaming, so it is not shown.
- Universal 3.6 Pro1.77% · 91ms
- Nemotron 3.5 Transcribe1.89%
NVIDIA Nemotron 3.5 Transcribe is ahead of AssemblyAI Universal 3.6 Pro 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
- Overall WER2.71%2.28%
- Pipecat Dataset WER1.77%1.89%
- Ocular Dataset WER4.50%3.03%
- Time to first text485msn/a
- Final-text delay91msn/a
Lower is better on every measure. Too close to call means the likely ranges overlap.
Universal 3.6 Pro vs Nemotron 3.5 Transcribe FAQ
Which is more accurate, AssemblyAI Universal 3.6 Pro or NVIDIA Nemotron 3.5 Transcribe?
NVIDIA Nemotron 3.5 Transcribe is more accurate on the public clips. On the 1,000 public Pipecat clips, AssemblyAI Universal 3.6 Pro scores 2.71% and NVIDIA Nemotron 3.5 Transcribe 2.28%, so NVIDIA Nemotron 3.5 Transcribe makes fewer errors by 0.43 points. On the eight licensed Ocular recordings, AssemblyAI Universal 3.6 Pro scores 1.77% and NVIDIA Nemotron 3.5 Transcribe 1.89%, with likely ranges of 1.55 to 2.00% and 1.63 to 2.17%. The ranges overlap, so the difference is too close to call. Word error rate counts incorrect, missing, and extra words, so lower is better.
NVIDIA Nemotron 3.5 Transcribe is more accurate on the public clips. On the 1,000 public Pipecat clips, AssemblyAI Universal 3.6 Pro scores 2.71% and NVIDIA Nemotron 3.5 Transcribe 2.28%, so NVIDIA Nemotron 3.5 Transcribe makes fewer errors by 0.43 points. On the eight licensed Ocular recordings, AssemblyAI Universal 3.6 Pro scores 1.77% and NVIDIA Nemotron 3.5 Transcribe 1.89%, with likely ranges of 1.55 to 2.00% and 1.63 to 2.17%. The ranges overlap, so the difference is too close to call. Word error rate counts incorrect, missing, and extra words, so lower is better.
They cannot be compared on speed. NVIDIA Nemotron 3.5 Transcribe was tested on whole recordings rather than a live stream, so its speed is not measured.
The two are too close to call on real conversations. NVIDIA Nemotron 3.5 Transcribe was tested on whole recordings rather than a live stream, so its speed is not measured. A voice agent cannot reply until the final transcript arrives, and the Ocular recordings are real conversations.
NVIDIA Nemotron 3.5 Transcribe was tested on whole recordings rather than a live stream, so whether it streams partial text was not measured. AssemblyAI Universal 3.6 Pro returns partial text before the speaker stops.
Test your voice agent the same way.
Cekura runs these simulated calls on your agent and scores every one.