Models listed
6
Published in this category
Worth weighing up
What goes in and out, licence terms, and whether you can host it yourself
How we treat data
Published specs only — we don't estimate numbers a vendor hasn't stated
Speaker diarization identifies who spoke when in an audio recording. It produces speaker-labeled time segments; it does not, by itself, transcribe the words or establish a person's real identity.
Test microphones, accents, short turns, background noise and overlapping speech. Measure missed speech, false speech and speaker confusion. When combining diarization with transcription, inspect the alignment between word timestamps and speaker turns.
Start with pyannote Community-1. Its profile explains access, preprocessing and the distinction between ordinary and exclusive diarization.
Scan the table below, then open any model for pricing, limits and the full write-up.
| Grok Voice Transcribe 2.0 xAI | — | — | — | — | — |
| Muse Voice Transcribe 1.0 Meta | — | — | — | — | — |
| Nemotron 3 Diarization NVIDIA Research | 10.0 | — | — | — | — |
| pyannote/speaker-diarization-community-1 pyannote | 5.0 | — | — | — | — |
| VibeVoice ASR Streaming 1.5B Microsoft Research | 5.0 | — | — | — | — |
| VibeVoice ASR Streaming 7B Microsoft Research | 5.0 | — | — | — | — |
1–6 of 6 models · select a column to sort
About this order: Recommended shows our editorial shortlist first, followed by recent releases. The spec score is a 0–100 figure SyncDev calculates from each vendor's published specifications — context window, reasoning support, input modalities, tool calling, maximum output and how recently the model shipped. It measures breadth of capability, not benchmark performance, so a higher-scoring model is not automatically the better choice for your task.