Downloads · 30 days
7
4% of all-time downloads
Onyx-Whisper-Tiny
wabibito/Onyx-Whisper-Tiny
Onyx-Whisper-Tiny is a automatic speech recognition model from wabibito. Use it when you need speech turned into text. It is set up for mlx. The card lists the license as apache-2.0.
MLX quantization of openai/whisper-tiny (Apache-2.0) for the Onyx on-device speech-to-text engine (OnyxSTT). Mixed precision: the AUDIO ENCODER is 4-bit (groupsize 64); the DECODER + embeddings + convs stay fp16. ~66…
All-time downloads
163
Public
Parameters
37.8M
65.8 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors65.8 MB · 94%
Parameter types
How the weights are stored.
F1630.7M · 81%
Base models
Model card
From the Hugging Face model README
Onyx-Whisper-Tiny (mixed 4-bit MLX)
MLX quantization of openai/whisper-tiny (Apache-2.0) for the Onyx on-device speech-to-text engine (OnyxSTT). Mixed precision: the AUDIO ENCODER is 4-bit (group_size 64); the DECODER + embeddings + convs stay fp16. ~66 MB (from ~151 MB fp32). Whispers autoregressive decoder is quant-sensitive (full 4-bit mishears numbers/codes), so only the robust encoder is 4-bit. Verified WER-neutral vs fp32 on clips with numbers/names/codes. Inherits the base Apache-2.0 license.
- Task
- Automatic Speech Recognition
- Library
- mlx
- Type
- whisper
- License
- apache-2.0
- Languages
- multilingual
- Created
- Jul 6, 2026
- Updated
- Jul 28, 2026