Vikhr-Llama-3.2-1B-instruct needs about 1.6 GB of memory at Q4_K_M (a 0.8 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 15 quants, from 0.6 GB (Q2_K) to 2.5 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.1 GB.
Compact Russian‑instruct Llama‑3.2 model for low‑power devices
Vikhr‑Llama‑3.2‑1B‑instruct is a 1.24B‑parameter Llama‑3.2 model fine‑tuned on the GrandMaster‑PRO‑MAX Russian dataset. It supports 131072‑token context and is optimized for mobile and low‑power deployment.
Good at
- Russian language instruction
- Chain‑of‑Thought support
- Low‑memory footprint
Limits
- No vision support
- License: llama3.2
- Requires 8k context for memory estimates
Best for
- Mobile chatbot
- Russian‑language tutoring
- Lightweight inference on edge devices
Who it’s for. Developers building Russian‑language assistants on devices with limited VRAM or RAM. Ideal for hobbyists and researchers needing a small, efficient instruct model.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 0.6 GB | 1.3 GB | Fits | Get file → | |
| IQ3_XS | 0.6 GB | 1.4 GB | Fits | Get file → | |
| Q3_K_S | 0.6 GB | 1.4 GB | Fits | Get file → | |
| IQ3_S | 0.6 GB | 1.4 GB | Fits | Get file → | |
| IQ3_M | 0.7 GB | 1.4 GB | Fits | Get file → | |
| Q3_K_M | 0.7 GB | 1.5 GB | Fits | Get file → | |
| Q3_K_L | 0.7 GB | 1.5 GB | Fits | Get file → | |
| IQ4_XS | 0.7 GB | 1.5 GB | Fits | Get file → | |
| Q4_K_S | 0.8 GB | 1.5 GB | Fits | Get file → | |
| Q4_K_M | 0.8 GB | 1.6 GB | Fits | Get file → | |
| Q5_K_S | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q5_K_M | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q6_K | 1.0 GB | 1.8 GB | Fits | Get file → | |
| Q8_0 | 1.3 GB | 2.1 GB | Fits | Get file → | |
| F16 | 2.5 GB | 3.2 GB | Fits | Get file → |
For 8 GB VRAM, use Q4_K_M (0.81 GB file, ~1.6 GB RAM). For 4 GB VRAM, choose Q3_K_S (0.64 GB file, ~1.4 GB RAM). For the smallest footprint, Q2_K (0.58 GB file, ~1.3 GB RAM) is suitable.
What GPU do you need for Vikhr-Llama-3.2-1B-instruct?
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 F16 (2.5 GB file, about 3.2 GB in use).
Which Vikhr-Llama-3.2-1B-instruct quant should you choose?
Q4_K_M is the usual starting point for Vikhr-Llama-3.2-1B-instruct: it balances file size and quality well. Here it is a 0.8 GB file.
The Q8_0 file is 1.3 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.
If memory is tight, the smallest file is Q2_K at 0.6 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Vikhr-Llama-3.2-1B-instruct
How much VRAM does Vikhr-Llama-3.2-1B-instruct need?
About 1.6 GB at Q4_K_M with an 8K context window. That is the 0.8 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 Vikhr-Llama-3.2-1B-instruct on an 8 GB GPU?
Yes, with F16: the 2.5 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 Vikhr-Llama-3.2-1B-instruct GGUF?
It depends on the quant: Q2_K is 0.6 GB, IQ3_XS is 0.6 GB, Q3_K_S is 0.6 GB, IQ3_S is 0.6 GB, IQ3_M is 0.7 GB, Q3_K_M is 0.7 GB and 9 more.
What context length does Vikhr-Llama-3.2-1B-instruct support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.3 GB at 8K and 1.1 GB at 32K.
Where can I download Vikhr-Llama-3.2-1B-instruct GGUF?
The files are on Hugging Face in the mradermacher/Vikhr-Llama-3.2-1B-Instruct-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 Vikhr-Llama-3.2-1B-instruct use?
Hugging Face lists the license as llama3.2. 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 Vikhr‑Llama‑3.2‑1B‑instruct on 8 GB VRAM?
Yes, the Q4_K_M quant uses about 1.6 GB RAM at 8k context, well within 8 GB VRAM.
Does the model support vision tasks?
No, the model is purely text‑based and has no vision capabilities.
What is the maximum context length?
The model was trained with a 131072‑token context window.