qwen3-tts-tokenizer-12hz needs about 0.8 GB of memory at Q8_0 (a 0.3 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 2 quants, from 0.3 GB (Q8_0) to 0.4 GB.
Qwen3‑TTS‑Tokenizer‑12Hz – lightweight speech tokenizer for 12.5 Hz encoding
The Qwen3‑TTS‑Tokenizer‑12Hz is a 0.162 billion‑parameter model that converts audio into discrete tokens. It uses a 16‑layer multi‑codebook design and a lightweight causal ConvNet to achieve extreme bitrate reduction and ultra‑low‑latency streaming.
Good at
- 12.5 Hz 16‑codebook speech representation
- Extreme bitrate reduction
- Ultra‑low‑latency streaming with first‑packet emission
Limits
- Only a tokenizer; requires a separate talker for synthesis
- No vision or multimodal support
- Requires 8 k context for memory estimates
Best for
- Encoding reference audio into ref_code
- Baking and loading voice‑pack GGUFs
- Decoding generated codes back to 24 kHz mono WAV
Who it’s for. Developers building TTS pipelines that need a compact, high‑fidelity tokenizer. It is intended to be paired with a Qwen3‑TTS talker model for full synthesis.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 0.3 GB | 0.8 GB | Fits | Get file → | |
| — | 0.4 GB | 0.9 GB | Fits | Get file → |
For 8 GB VRAM, the Q8_0 quantised file (0.29 GB, 0.8 GB memory at 8 k context) is recommended. If more precision is needed and memory allows, use the unknown quantised file (0.36 GB, 0.9 GB memory). The F16 baseline is 0.29 GB but requires more memory than the Q8_0 variant.
What GPU do you need for qwen3-tts-tokenizer-12hz?
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 (0.4 GB file, about 0.9 GB in use).
Which qwen3-tts-tokenizer-12hz quant should you choose?
Q8_0 (0.3 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about qwen3-tts-tokenizer-12hz
How much VRAM does qwen3-tts-tokenizer-12hz need?
About 0.8 GB at Q8_0 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 qwen3-tts-tokenizer-12hz on an 8 GB GPU?
Yes, with the smaller files: the 0.4 GB file needs about 0.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of qwen3-tts-tokenizer-12hz GGUF?
It depends on the quant: Q8_0 is 0.3 GB, file is 0.4 GB.
Where can I download qwen3-tts-tokenizer-12hz GGUF?
The files are on Hugging Face in the cstr/qwen3-tts-tokenizer-12hz-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 qwen3-tts-tokenizer-12hz 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 this tokenizer on 8 GB VRAM?
Yes, the Q8_0 variant uses 0.8 GB memory at 8 k context, well within an 8 GB VRAM budget.
Does this tokenizer support multiple languages?
Yes, it supports ten languages: zh, en, ja, ko, de, fr, ru, pt, es, it.
Do I need a separate model for speech synthesis?
Yes, this tokenizer is only for encoding/decoding. Pair it with a Qwen3‑TTS talker for full synthesis.