AssemblyAI Universal 3.5 Pro vs GPT Realtime Whisper
AssemblyAI Universal 3.5 Pro is ahead on the Pipecat Dataset, time to first text and final-text delay, and GPT Realtime Whisper on the Ocular Dataset.
Quality and speed frontier
Every model's error rate on the public clips, against how long its transcript takes to settle.
The four metrics, side by side
Word error rate on 1,000 clips from Pipecat's STT benchmark dataset.
- AssemblyAI Universal 3.5 Pro
- 1.93%1st
- GPT Realtime Whisper
- 2.16%4th
AssemblyAI Universal 3.5 Pro, 0.23 points lower
Word error rate on eight licensed recordings provided by Ocular.
- AssemblyAI Universal 3.5 Pro
- 4.23%9th
- GPT Realtime Whisper
- 3.53%4th
GPT Realtime Whisper, 0.70 points lower
Median delay before the first words arrive, from the first audio packet.
- AssemblyAI Universal 3.5 Pro
- 489ms1st
- GPT Realtime Whisper
- 1.38s8th
AssemblyAI Universal 3.5 Pro, 890ms sooner
Median delay after the speaker stops before the transcript is final.
- AssemblyAI Universal 3.5 Pro
- 180ms7th
- GPT Realtime Whisper
- 543ms12th
AssemblyAI Universal 3.5 Pro, 363ms sooner
Lower is better on every metric. The small figure is each model's place in the whole field.
First text and final text, at three percentiles
Both measured on the same 206 streamed Ocular Dataset turns. P90 and P95 say what a caller meets on a bad turn, which a median hides.
| Measure | Model | P50 | P90 | P95 |
|---|---|---|---|---|
| Time to first text | AssemblyAI Universal 3.5 Pro | 489ms | 966ms | 1.08s |
| GPT Realtime Whisper | 1.38s | 1.84s | 2.86s | |
| Final-text delay | AssemblyAI Universal 3.5 Pro | 180ms | 532ms | 692ms |
| GPT Realtime Whisper | 543ms | 647ms | 669ms |
Both against every other model
Open on the metric where the two sit furthest apart, with the other three a tab away.
Time to first text, measured from the first audio packet. Lower is better.
| # | Model | First text | turns | |
|---|---|---|---|---|
| 1 | AssemblyAI Universal 3.5 Pro | 489ms | 205 | |
| 2 | Deepgram Flux English | 791ms | 199 | |
| 3 | Deepgram Flux Multilingual | 793ms | 175 | |
| 4 | Smallest Pulse | 1057ms | 206 | |
| 5 | Inworld STT-1 | 1063ms | 206 | |
| 6 | Deepgram Nova-3 | 1064ms | 203 | |
| 7 | Speechmatics Linden | 1190ms | 194 | |
| 8 | GPT Realtime Whisper | 1379ms | 206 | |
| 9 | Cartesia Ink 2 | 1511ms | 198 | |
| 10 | Gradium | 1705ms | 198 | |
| 11 | Reson8 | 2400ms | 201 | |
| 12 | Google Chirp 3 | 4480ms | 206 | |
| 13 | Google Chirp 2 | 4497ms | 205 |
Not ranked: GPT-4o Mini Transcribe and GPT-4o Transcribe return no text until the speaker stops, so first-text time is not comparable.
Median across the same 206 Ocular Dataset turns. Providers finalize differently, so compare models with similar finalization contracts before treating small gaps as meaningful.
Both on the other two views
Pipecat vs Ocular accuracy
Every model's error rate on the public clips, against the same model on real conversations.
Transcription response time
How soon each model returns its first words, against how soon its transcript is final.
What the numbers rest on
| Model | Pipecat clips | Ocular recordings | Turns timed | Failed |
|---|---|---|---|---|
| AssemblyAI Universal 3.5 Pro | 1,000 | 8 | 205 | 1 |
| GPT Realtime Whisper | 1,000 | 8 | 206 | 0 |
A clip counts only where the model returned usable text, so counts differ between models. Word error rate is measured the same way throughout: incorrect, missing, and extra words against the reference transcript.
Frequently asked questions
Is AssemblyAI Universal 3.5 Pro or GPT Realtime Whisper more accurate?
On the 1,000 public Pipecat clips, AssemblyAI Universal 3.5 Pro scores 1.93% and GPT Realtime Whisper 2.16%, so AssemblyAI Universal 3.5 Pro makes fewer errors by 0.23 points. On the eight licensed Ocular recordings, AssemblyAI Universal 3.5 Pro scores 4.23% and GPT Realtime Whisper 3.53%, so GPT Realtime Whisper makes fewer errors by 0.70 points. The two datasets disagree, so the better choice depends on whether your audio looks more like clean public clips or real conversations.
Which is faster, AssemblyAI Universal 3.5 Pro or GPT Realtime Whisper?
AssemblyAI Universal 3.5 Pro returns first text 890ms sooner, at 489ms against 1.38s at the median. AssemblyAI Universal 3.5 Pro has the final transcript 363ms sooner, 180ms after the speaker stops against 543ms. Both are measured on the same 206 streamed turns.
How were AssemblyAI Universal 3.5 Pro and GPT Realtime Whisper compared?
Both ran the same two datasets under the same conditions as every other model in the benchmark: 1,000 public clips from Pipecat's STT benchmark dataset and eight licensed conversation recordings from Ocular. AssemblyAI Universal 3.5 Pro returned usable text for 1,000 of the public clips and GPT Realtime Whisper for 1,000. Latency uses each provider's own supported finalization contract.