kokoro needs about 0.7 GB of memory at its main file (a 0.2 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 10 quants, from 0.2 GB to 0.4 GB.
Kokoro: lightweight 82M parameter TTS model
Kokoro is an open-weight text‑to‑speech model with 82 million parameters. It delivers quality comparable to larger models while remaining fast and cost‑efficient. The model is licensed under MIT and can be deployed in production or personal projects.
Good at
- Fast inference on modest hardware
- High‑quality speech synthesis
- Low memory footprint
Limits
- No vision support
- Context length not specified
- Requires espeak or native phonemization
Best for
- Embedded TTS on edge devices
- Rapid prototyping of voice assistants
- Low‑cost audio generation
Who it’s for. Developers and researchers needing a small, efficient TTS model that can run on limited GPU or CPU resources. The MIT license allows unrestricted use in commercial or open‑source projects.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| F16 | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| F16 | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.4 GB | 0.9 GB | Fits | Get file → | |
| — | 0.4 GB | 0.9 GB | Fits | Get file → |
For 8 GB VRAM, the F16 quantized file (0.2 GB) uses about 0.7 GB memory at 8 k context, which is well within budget. If memory is even tighter, choose a smaller quant file such as the 0.18 GB unknown quant, which also estimates 0.7 GB usage. All estimates are approximate.
What GPU do you need for kokoro?
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 kokoro quant should you choose?
If memory is tight, the smallest file is the first one in the table at 0.2 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about kokoro
How much VRAM does kokoro need?
About 0.7 GB at its main file with an 8K context window. That is the 0.2 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 kokoro 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 kokoro GGUF?
It depends on the quant: file is 0.2 GB, file is 0.2 GB, file is 0.2 GB, file is 0.2 GB, file is 0.2 GB, F16 is 0.2 GB and 4 more.
Where can I download kokoro GGUF?
The files are on Hugging Face in the mmwillet2/Kokoro_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 kokoro use?
Hugging Face lists the license as mit. 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 Kokoro on 4 GB VRAM?
The smallest available quantized file is 0.18 GB, but it still estimates 0.7 GB memory at 8 k context, so 4 GB VRAM is sufficient for inference.
Does Kokoro support multi‑language voices?
The model does not list supported languages; it is primarily a single‑language TTS model.
Is Kokoro compatible with llama.cpp?
Yes, Kokoro is distributed as GGUF files compatible with tts.cpp, which can be used with llama.cpp for inference.