multitalker parakeet streaming 0.6b v1 needs about 1.2 GB of memory at Q4_K_M (a 0.6 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 6 quants, from 0.6 GB (Q4_K_M) to 3.0 GB (F32).
Multitalker Parakeet Streaming 0.6B v1 – real‑time speaker‑aware ASR
Multitalker Parakeet Streaming 0.6B v1 is a streaming ASR model that requires one instance per speaker. It eliminates speaker enrollment by using speaker activity from diarization and adapts dynamically via learnable speaker kernels. The model supports up to four overlapping speakers with speaker‑tagged transcripts.
Good at
- Real‑time streaming
- Speaker‑aware transcription
- Handles fully overlapped speech
Limits
- Requires one model instance per speaker
- No explicit speaker enrollment
- No vision support
Best for
- Live conference transcription
- Multispeaker podcast editing
- Real‑time customer support calls
Who it’s for. Developers building real‑time speech‑to‑text applications that need accurate transcription of multiple speakers without prior enrollment. Ideal for systems that can allocate separate model instances per speaker.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 0.6 GB | 1.2 GB | Fits | Get file → | |
| Q5_K_M | 0.7 GB | 1.3 GB | Fits | Get file → | |
| Q6_K | 0.7 GB | 1.3 GB | Fits | Get file → | |
| Q8_0 | 0.9 GB | 1.5 GB | Fits | Get file → | |
| F16 | 1.5 GB | 2.2 GB | Fits | Get file → | |
| F32 | 3.0 GB | 3.8 GB | Fits | Get file → |
For 8 GB VRAM, use Q4_K_M (0.62 GB, ~1.2 GB memory at 8k context) or Q5_K_M (0.68 GB, ~1.3 GB). For 16 GB VRAM, Q6_K (0.74 GB, ~1.3 GB) or Q8_0 (0.87 GB, ~1.5 GB) are suitable. Higher precision F16 (1.48 GB) or F32 (2.96 GB) requires larger memory budgets.
What GPU do you need for multitalker parakeet streaming 0.6b v1?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is F32 (3.0 GB file, about 3.8 GB in use).
Which multitalker parakeet streaming 0.6b v1 quant should you choose?
Q4_K_M is the usual starting point for multitalker parakeet streaming 0.6b v1: it balances file size and quality well. Here it is a 0.6 GB file.
The Q8_0 file is 0.9 GB, about 1.4 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about multitalker parakeet streaming 0.6b v1
How much VRAM does multitalker parakeet streaming 0.6b v1 need?
About 1.2 GB at Q4_K_M with an 8K context window. That is the 0.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run multitalker parakeet streaming 0.6b v1 on an 8 GB GPU?
Yes, with F32: the 3.0 GB file needs about 3.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of multitalker parakeet streaming 0.6b v1 GGUF?
It depends on the quant: Q4_K_M is 0.6 GB, Q5_K_M is 0.7 GB, Q6_K is 0.7 GB, Q8_0 is 0.9 GB, F16 is 1.5 GB, F32 is 3.0 GB.
Where can I download multitalker parakeet streaming 0.6b v1 GGUF?
The files are on Hugging Face in the handy-computer/multitalker-parakeet-streaming-0.6b-v1-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does multitalker parakeet streaming 0.6b v1 use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run this model on 8 GB VRAM?
Yes, using Q4_K_M or Q5_K_M quantizations keeps memory usage below 1.5 GB at 8k context.
Does this model support more than four speakers?
The bundled GGUFs support up to four speakers; additional speakers would need separate instances.
Is speaker enrollment required?
No, the model uses diarization outputs and learns speaker kernels dynamically.