TranscribePsy 250M EU sub needs about 0.8 GB of memory at its main file (a 0.3 GB file) with an 8K context window, so it fits on a 4 GB GPU.
Compact 250M-parameter speech recognition for eight European languages
TranscribePsy-250M-EU-sub is a 254M-parameter model built for on-device speech recognition. It covers Dutch, English, French, German, Italian, Polish, Portuguese, and Spanish. The model is distilled from Parakeet-TDT-0.6B-v3 and optimized for edge deployment.
Good at
- Accurate on eight European languages
- Fast real‑time factor 0.052
- Low inference latency 1,021 ms
Limits
- No vision support
- Limited to 8‑k context
- Requires 16 kHz mono WAV
Best for
- Mobile transcription apps
- Edge speech-to-text services
- Low‑latency voice assistants
Who it’s for. Developers building on‑device speech recognition for the eight supported European languages. Ideal for mobile and embedded systems that need quick, accurate transcription without large models.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.3 GB | 0.8 GB | Fits | Get file → |
The available GGUF build is 0.3 GB and estimates 0.8 GB memory at 8 k context. For devices with 1 GB VRAM, use the Q8_0 quantized build. If memory is tighter, consider a lower‑bit quantization if available, but note the file list only includes the Q8_0 build.
What GPU do you need for TranscribePsy 250M EU sub?
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 the main file (0.3 GB file, about 0.8 GB in use).
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about TranscribePsy 250M EU sub
How much VRAM does TranscribePsy 250M EU sub need?
About 0.8 GB at its main file with an 8K context window. That is the 0.3 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 TranscribePsy 250M EU sub on an 8 GB GPU?
Yes, with the smaller files: the 0.3 GB file needs about 0.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of TranscribePsy 250M EU sub GGUF?
It depends on the quant: file is 0.3 GB.
Where can I download TranscribePsy 250M EU sub GGUF?
The files are on Hugging Face in the qvac/TranscribePsy-250M-EU-sub-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 TranscribePsy 250M EU sub use?
Hugging Face lists the license as apache-2.0. 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 TranscribePsy-250M-EU-sub on a phone with 8 GB VRAM?
Yes. The 0.3 GB GGUF file uses about 0.8 GB memory at 8 k context, well within 8 GB VRAM.
Does the model support languages other than the eight listed?
No. The model only supports Dutch, English, French, German, Italian, Polish, Portuguese, and Spanish.
What audio format does the engine accept?
The engine requires 16 kHz mono WAV files; it does not resample audio.