Sarvamai.Sarvam 30b needs about 20.2 GB of memory at F16 (a 19.6 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 6 quants, from 12.2 GB (F16) to 34.2 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 0.6 GB.
Sarvamai.Sarvam 30b – a 32.2B‑parameter MoE model for efficient inference
Sarvamai.Sarvam 30b is a 32.2B‑parameter mixture‑of‑experts model based on the Sarvam architecture. It supports a 131072‑token context window and is released under the Apache‑2.0 license. The model is optimized for throughput and memory efficiency with 19 layers and 128 experts.
Good at
- Long‑context reasoning
- Multilingual Indian language support
- Efficient inference on limited GPU memory
Limits
- Requires 8‑bit or higher precision for full performance
- No vision or audio modalities
- Large VRAM needed for full‑precision inference
Best for
- Deploying conversational agents in Indian languages
- Running long‑form content generation
- Tool‑calling workflows in resource‑constrained environments
Who it’s for. Developers building multilingual LLM services that need strong reasoning and coding capabilities while keeping GPU memory usage low. The model is suitable for production deployments where throughput and efficiency are critical.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| F16 | 12.2 GB | 12.8 GB | Too big | Get file → | |
| F16 | 15.7 GB | 16.3 GB | Too big | Get file → | |
| F16 | 19.6 GB | 20.2 GB | Too big | Get file → | |
| F16 | 22.9 GB | 23.6 GB | Too big | Get file → | |
| F16 | 26.4 GB | 27.1 GB | Too big | Get file → | |
| F16 | 34.2 GB | 34.9 GB | Too big | Get file → |
For 8 GB VRAM, use the 12.16 GB F16 file (≈12.8 GB memory at 8 k context). For 12 GB VRAM, the 15.68 GB F16 file (≈16.3 GB memory) is suitable. Larger VRAM allows the 19.58 GB or 22.92 GB variants for higher throughput.
What GPU do you need for Sarvamai.Sarvam 30b?
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 F16 (12.2 GB file, about 12.8 GB in use).
- 24 GB VRAM: the best quant that fits is F16 (22.9 GB file, about 23.6 GB in use).
- 32 GB VRAM: the best quant that fits is F16 (26.4 GB file, about 27.1 GB in use).
- 48 GB VRAM: the best quant that fits is F16 (34.2 GB file, about 34.9 GB in use).
Which Sarvamai.Sarvam 30b quant should you choose?
F16 (19.6 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
If memory is tight, the smallest file is F16 at 12.2 GB, with the quality trade-offs that come with fewer bits. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Sarvamai.Sarvam 30b
How much VRAM does Sarvamai.Sarvam 30b need?
About 20.2 GB at F16 with an 8K context window. That is the 19.6 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 Sarvamai.Sarvam 30b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 12.8 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Sarvamai.Sarvam 30b GGUF?
It depends on the quant: F16 is 12.2 GB, F16 is 15.7 GB, F16 is 19.6 GB, F16 is 22.9 GB, F16 is 26.4 GB, F16 is 34.2 GB.
What context length does Sarvamai.Sarvam 30b support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.6 GB at 32K.
Where can I download Sarvamai.Sarvam 30b GGUF?
The files are on Hugging Face in the DevQuasar/sarvamai.sarvam-30b-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 Sarvamai.Sarvam 30b 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 Sarvamai.Sarvam 30b on 8 GB VRAM?
Use the 12.16 GB F16 quantized file; it requires about 12.8 GB memory at 8 k context, which is close to 8 GB VRAM but may need swapping.
Does the model support vision or audio?
No, the model only provides text generation capabilities.
What license applies to this model?
The model is released under the Apache‑2.0 license.