FireRedASR2-AED needs about 2.1 GB of memory at Q8_0 (a 1.5 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 3 quants, from 1.0 GB (Q4_K) to 2.4 GB.
FireRedASR2-AED: 1.18B parameter ASR model in GGUF format
FireRedASR2-AED is a 1.18B parameter Automatic Speech Recognition model. It supports Mandarin Chinese, English, and 20+ Chinese dialects. The model is released under Apache 2.0 and is available in multiple GGUF quantisations.
Good at
- Mandarin and English transcription
- Supports 20+ Chinese dialects
- Low‑memory GGUF variants
Limits
- No vision support
- Requires at least 8k context for memory estimates
- No beam search decoding
Best for
- Transcribing Mandarin speech
- Transcribing code‑switching Mandarin/English
- Deploying on edge devices with limited VRAM
Who it’s for. Developers building speech‑to‑text pipelines for Mandarin and English. Researchers needing a lightweight ASR model with open licence. Engineers deploying ASR on devices with limited GPU memory.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K | 1.0 GB | 1.6 GB | Fits | Get file → | |
| Q8_0 | 1.5 GB | 2.1 GB | Fits | Get file → | |
| — | 2.4 GB | 3.2 GB | Fits | Get file → |
For 8 GB VRAM, use the Q4_K variant (0.96 GB file, 1.6 GB estimated memory). If 16 GB VRAM is available, the Q8_0 variant (1.46 GB file, 2.1 GB estimated memory) offers higher precision. The unknown quant is 2.41 GB and 3.2 GB memory, suitable for larger GPU setups.
What GPU do you need for FireRedASR2-AED?
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 (2.4 GB file, about 3.2 GB in use).
Which FireRedASR2-AED quant should you choose?
Q8_0 (1.5 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.
If memory is tight, the smallest file is Q4_K at 1.0 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about FireRedASR2-AED
How much VRAM does FireRedASR2-AED need?
About 2.1 GB at Q8_0 with an 8K context window. That is the 1.5 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 FireRedASR2-AED on an 8 GB GPU?
Yes, with the smaller files: the 2.4 GB file needs about 3.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of FireRedASR2-AED GGUF?
It depends on the quant: Q4_K is 1.0 GB, Q8_0 is 1.5 GB, file is 2.4 GB.
Where can I download FireRedASR2-AED GGUF?
The files are on Hugging Face in the cstr/firered-asr2-aed-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 FireRedASR2-AED 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 FireRedASR2-AED on 8 GB VRAM?
Yes, the Q4_K variant requires about 1.6 GB of VRAM at 8k context, fitting comfortably in 8 GB.
Does the model support beam search decoding?
No, only CTC decoding is implemented; beam search is not yet available.
What license applies to the GGUF files?
The files are distributed under Apache 2.0, the same licence as the upstream model.