GigaChat 20B A3B instruct v1.5 needs about 13.9 GB of memory at Q4_K_M (a 12.5 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 7 quants, from 11.7 GB (Q4_0) to 41.2 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.8 GB.
GigaChat 20B A3B instruct v1.5 – 20.6B parameter dialog model
GigaChat 20B A3B instruct v1.5 is a 20.6B parameter dialog model based on deepseek architecture. It supports a 131 072 token context window and is released under the MIT license. The model is available in multiple GGUF quantizations.
Good at
- Large context handling
- High alignment scores
- Multi‑expert architecture
Limits
- No vision support
- Requires 8 k context for memory estimates
- No explicit language support list
Best for
- Chat applications needing long context
- Fine‑tuning with custom prompts
- Research on mixture‑of‑experts models
Who it’s for. Developers building chat systems that require extended context and high alignment. Researchers interested in mixture‑of‑experts architectures and large‑scale dialog models.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 11.7 GB | 13.1 GB | Too big | Get file → | |
| Q4_K_M | 12.5 GB | 13.9 GB | Too big | Get file → | |
| Q5_0 | 14.2 GB | 15.6 GB | Too big | Get file → | |
| Q5_K_M | 14.6 GB | 16.1 GB | Too big | Get file → | |
| Q6_K | 16.9 GB | 18.3 GB | Too big | Get file → | |
| Q8_0 | 21.9 GB | 23.3 GB | Too big | Get file → | |
| BF16 | 41.2 GB | 42.6 GB | Too big | Get file → |
For 8 GB VRAM, Q4_0 (file size 11.67 GB, estimated 13.1 GB memory) is suitable. If 12 GB VRAM is available, Q4_K_M (12.49 GB, 13.9 GB memory) can be used. Higher precision quants like Q6_K or BF16 require more memory and are not recommended for limited VRAM setups.
What GPU do you need for GigaChat 20B A3B instruct v1.5?
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: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: the best quant that fits is Q5_0 (14.2 GB file, about 15.6 GB in use).
- 24 GB VRAM: the best quant that fits is Q8_0 (21.9 GB file, about 23.3 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (21.9 GB file, about 23.3 GB in use).
- 48 GB VRAM: the best quant that fits is BF16 (41.2 GB file, about 42.6 GB in use).
Which GigaChat 20B A3B instruct v1.5 quant should you choose?
Q4_K_M is the usual starting point for GigaChat 20B A3B instruct v1.5: it balances file size and quality well. Here it is a 12.5 GB file.
The Q8_0 file is 21.9 GB, about 1.8 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_0 at 11.7 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 GigaChat 20B A3B instruct v1.5
How much VRAM does GigaChat 20B A3B instruct v1.5 need?
About 13.9 GB at Q4_K_M with an 8K context window. That is the 12.5 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 GigaChat 20B A3B instruct v1.5 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 13.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of GigaChat 20B A3B instruct v1.5 GGUF?
It depends on the quant: Q4_0 is 11.7 GB, Q4_K_M is 12.5 GB, Q5_0 is 14.2 GB, Q5_K_M is 14.6 GB, Q6_K is 16.9 GB, Q8_0 is 21.9 GB and 1 more.
What context length does GigaChat 20B A3B instruct v1.5 support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.9 GB at 8K and 3.8 GB at 32K.
Where can I download GigaChat 20B A3B instruct v1.5 GGUF?
The files are on Hugging Face in the ai-sage/GigaChat-20B-A3B-instruct-v1.5-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 GigaChat 20B A3B instruct v1.5 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 GigaChat 20B A3B instruct on 8 GB VRAM?
Yes, the Q4_0 quantization uses 13.1 GB estimated memory at 8 k context, which fits within 8 GB VRAM with some overhead.
Does the model support vision tasks?
No, the model does not include vision capabilities.
What license applies to this model?
The model is released under the MIT license.