Jais ASR needs about 4.9 GB of memory at Q4_K_M (a 3.2 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 5 quants, from 2.9 GB (Q4_K_S) to 5.4 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.8 GB.
Jais ASR – Arabic speech recognition with code‑switching
Jais ASR is a 5.06B‑parameter model for Arabic and dialects, supporting code‑switching with English, French, Spanish, and Italian. It includes optional diacritization, language and dialect detection, and a no‑speech signal. The model is released under Apache‑2.0 and available in GGUF format.
Good at
- Arabic ASR with low WER
- Code‑switching detection
- Optional diacritization
Limits
- No vision support
- Requires MBZUAI‑IFM fork of llama.cpp
- Large context length 524288 tokens
Best for
- Transcribing Arabic podcasts
- Multilingual customer support in Arabic
- Real‑time Arabic speech analytics
Who it’s for. Developers building Arabic speech‑to‑text services or integrating Arabic ASR into multilingual applications. The model is suitable for production use with vLLM or llama.cpp.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_S | 2.9 GB | 4.6 GB | Fits | Get file → | |
| IQ4_XS | 2.9 GB | 4.6 GB | Fits | Get file → | |
| Q4_K_M | 3.2 GB | 4.9 GB | Fits | Get file → | |
| Q5_K_M | 3.6 GB | 5.4 GB | Fits | Get file → | |
| Q8_0 | 5.4 GB | 7.1 GB | Fits | Get file → |
For 8 GB VRAM, use Q4_K_S or IQ4_XS (≈4.6 GB memory). For 12 GB VRAM, Q4_K_M or Q5_K_M (≈4.9–5.4 GB). For 16 GB VRAM, Q8_0 (≈7.1 GB). Memory estimates are approximate.
What GPU do you need for Jais ASR?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) 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: no quant fits at an 8K context.
- 6 GB VRAM: the best quant that fits is Q5_K_M (3.6 GB file, about 5.4 GB in use).
- 8 GB VRAM: the best quant that fits is Q8_0 (5.4 GB file, about 7.1 GB in use).
Which Jais ASR quant should you choose?
Q4_K_M is the usual starting point for Jais ASR: it balances file size and quality well. Here it is a 3.2 GB file.
The Q8_0 file is 5.4 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q4_K_S at 2.9 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 Jais ASR
How much VRAM does Jais ASR need?
About 4.9 GB at Q4_K_M with an 8K context window. That is the 3.2 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 Jais ASR on an 8 GB GPU?
Yes, with Q8_0: the 5.4 GB file needs about 7.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Jais ASR GGUF?
It depends on the quant: Q4_K_S is 2.9 GB, IQ4_XS is 2.9 GB, Q4_K_M is 3.2 GB, Q5_K_M is 3.6 GB, Q8_0 is 5.4 GB.
What context length does Jais ASR support?
The GGUF header lists a trained context length of 524,288 tokens (512K). The KV cache grows with the context you use: about 1.2 GB at 8K and 4.8 GB at 32K.
Where can I download Jais ASR GGUF?
The files are on Hugging Face in the IFM/Jais-ASR-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 Jais ASR 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 Jais ASR on 8 GB VRAM?
Yes, the Q4_K_S or IQ4_XS quantizations use about 4.6 GB of GPU memory at 8 k context.
Does Jais ASR support vision?
No, the model only handles audio and text decoding.
What license applies to Jais ASR?
The model is released under the Apache‑2.0 license.