GigaChat3 10B A1.8B Bf16 needs about 7.0 GB of memory at its main file (a 6.1 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 5 quants, from 3.3 GB to 11.4 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 1.9 GB.
GigaChat3 10B A1.8B – a Mixture‑of‑Experts dialogue model
GigaChat3 10B A1.8B is a 10.7B parameter model with 1.8B active experts. It supports up to 262144 context tokens and is released under the MIT license. The model is optimized for high throughput inference with Multi‑head Latent Attention and Multi‑Token Prediction.
Good at
- Large context handling
- High inference speed
- Mixture‑of‑Experts efficiency
Limits
- No vision support
- Requires specialized quant for full size
- Large VRAM needed for full precision
Best for
- Chatbots needing long context
- Research prototyping with MoE
- High‑throughput inference pipelines
Who it’s for. Developers and researchers who need a dialogue model with extended context and efficient inference. The model is suitable for applications that can leverage its MoE architecture and high token limits.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 3.3 GB | 4.2 GB | Fits | Get file → | |
| — | 4.2 GB | 5.1 GB | Fits | Get file → | |
| — | 6.1 GB | 7.0 GB | Fits | Get file → | |
| — | 8.2 GB | 9.1 GB | Too big | Get file → | |
| Q8_0 | 11.4 GB | 12.3 GB | Too big | Get file → |
For 8 GB VRAM, use the Q8_0 quant (11.35 GB file, ~12.3 GB memory at 8 k context). For 12 GB VRAM, the unknown 8.16 GB quant (~9.1 GB memory) is a good fit. Larger VRAM allows the 6.08 GB quant (~7 GB memory).
What GPU do you need for GigaChat3 10B A1.8B Bf16?
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 the main file (4.2 GB file, about 5.1 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (6.1 GB file, about 7.0 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (8.2 GB file, about 9.1 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (11.4 GB file, about 12.3 GB in use).
Which GigaChat3 10B A1.8B Bf16 quant should you choose?
If memory is tight, the smallest file is the first one in the table at 3.3 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about GigaChat3 10B A1.8B Bf16
How much VRAM does GigaChat3 10B A1.8B Bf16 need?
About 7.0 GB at its main file with an 8K context window. That is the 6.1 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 GigaChat3 10B A1.8B Bf16 on an 8 GB GPU?
Yes, with the smaller files: the 6.1 GB file needs about 7.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of GigaChat3 10B A1.8B Bf16 GGUF?
It depends on the quant: file is 3.3 GB, file is 4.2 GB, file is 6.1 GB, file is 8.2 GB, Q8_0 is 11.4 GB.
What context length does GigaChat3 10B A1.8B Bf16 support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.5 GB at 8K and 1.9 GB at 32K.
Where can I download GigaChat3 10B A1.8B Bf16 GGUF?
The files are on Hugging Face in the ubergarm/GigaChat3-10B-A1.8B-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 GigaChat3 10B A1.8B Bf16 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 GigaChat3 10B A1.8B on 8 GB VRAM?
Only the Q8_0 quant can be used, as it requires about 12.3 GB memory at 8 k context.
Does the model support image input?
No, the model has no vision capabilities.
What license applies to this model?
The model is released under the MIT license.