The AI can distinguish between different speakers and label them throughout the transcript. How to Install Speech to Text v12.0
| Feature | Speech to Text v11 (2022) | | | :--- | :--- | :--- | | Time to transcribe (1 hour, 8 tracks) | 12 minutes | 4 minutes | | Speaker ID accuracy (2 speakers) | 78% | 94% | | Punctuation accuracy | Fair (misses question marks) | Excellent (contextual commas) | | Memory footprint | 1.2 GB | 800 MB (optimized) |
: