MERaLiON 2 3B needs about 3.1 GB of memory at Q4_K_M (a 1.7 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 2 quants, from 1.7 GB (Q4_K_M) to 2.8 GB (Q8_0).
MERaLiON 2 3B – 2.61B parameter Gemma2 model with vision support
MERaLiON 2 3B is a 2.61B parameter Gemma2 model designed for speech‑text tasks. It includes vision capabilities and supports up to 8192 context tokens. The model is available in Q4_K_M and Q8_0 quantizations.
Good at
- Speech‑text multimodal tasks
- Vision‑enabled inference
- Low‑latency inference on modest hardware
Limits
- Requires mmproj file for decoder
- No vision support without mmproj
- License is other, not specified
Best for
- Audio‑question answering up to 5 minutes
- Speech transcription and translation
- Multilingual dialogue with vision prompts
Who it’s for. Developers building lightweight speech‑text applications that need vision support. Ideal for teams with limited GPU memory but requiring multimodal inference.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 1.7 GB | 3.1 GB | Fits | Get file → | |
| Q8_0 | 2.8 GB | 4.2 GB | Fits | Get file → |
For 8 GB VRAM use Q4_K_M (1.71 GB, ~3.1 GB memory at 8k context). For 12 GB VRAM use Q8_0 (2.78 GB, ~4.2 GB memory at 8k context). Both need the mmproj file (~1.7 GB).
What GPU do you need for MERaLiON 2 3B?
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: the best quant that fits is Q4_K_M (1.7 GB file, about 3.1 GB in use).
- 6 GB VRAM: the best quant that fits is Q8_0 (2.8 GB file, about 4.2 GB in use).
Which MERaLiON 2 3B quant should you choose?
Q4_K_M is the usual starting point for MERaLiON 2 3B: it balances file size and quality well. Here it is a 1.7 GB file.
The Q8_0 file is 2.8 GB, about 1.6 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.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about MERaLiON 2 3B
How much VRAM does MERaLiON 2 3B need?
About 3.1 GB at Q4_K_M with an 8K context window. That is the 1.7 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 MERaLiON 2 3B on an 8 GB GPU?
Yes, with Q8_0: the 2.8 GB file needs about 4.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of MERaLiON 2 3B GGUF?
It depends on the quant: Q4_K_M is 1.7 GB, Q8_0 is 2.8 GB.
What context length does MERaLiON 2 3B support?
The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 0.9 GB at 8K and 0.9 GB at 8K.
Where can I download MERaLiON 2 3B GGUF?
The files are on Hugging Face in the MERaLiON/MERaLiON-2-3B-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 MERaLiON 2 3B use?
Hugging Face lists the license as other. 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 MERaLiON 2 3B on 8 GB VRAM?
Yes, the Q4_K_M quantization is 1.71 GB and requires about 3.1 GB memory at 8k context, fitting within 8 GB VRAM.
Do I need a separate mmproj file?
Yes, both decoder files require the mmproj file (~1.7 GB) for vision support.
Is the model suitable for speech‑to‑text only?
The model includes vision support but can be used for speech‑to‑text tasks; however, the mmproj file is still required for full functionality.