SA

Sarvam Translate

SarvamGGUF by mradermacherUpdated 2026-10-13
Download on Hugging Face
Parameters3.88B
Context128K
Layers34
Architecturegemma3
Smallest quant1.7 GB
Quick answer

Sarvam Translate needs about 4.1 GB of memory at Q4_K_M (a 2.5 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 12 quants, from 1.7 GB (Q2_K) to 7.8 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 4.6 GB.

Sarvam Translate – document‑level translation for 22 Indian languages

Sarvam Translate is a 3.88B‑parameter Gemma3 model designed for comprehensive document‑level translation across 22 official Indian languages. It supports long‑context inputs up to 8k tokens and handles diverse formats such as markdown, OCR‑text, math, chemistry, and code comments.

Good at

  • Long‑context Indian language translation
  • Document‑level multi‑format handling
  • Context‑aware style and inclusivity

Limits

  • No vision support beyond detected abilities
  • GPL‑3.0 license restricts commercial use
  • Maximum context length 8k tokens

Best for

  • Translating legal documents
  • Academic papers in Indian languages
  • Multilingual web content

Who it’s for. Developers and researchers needing high‑quality translation for Indian languages in document‑level contexts. The model is open‑source under GPL‑3.0, suitable for academic and non‑commercial projects.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K1.7 GB 3.4 GB
Fits Get file →
Q3_K_S1.9 GB 3.6 GB
Fits Get file →
Q3_K_M2.1 GB 3.7 GB
Fits Get file →
Q3_K_L2.2 GB 3.9 GB
Fits Get file →
IQ4_XS2.3 GB 3.9 GB
Fits Get file →
Q4_K_S2.4 GB 4.0 GB
Fits Get file →
Q4_K_M2.5 GB 4.1 GB
Fits Get file →
Q5_K_S2.8 GB 4.4 GB
Fits Get file →
Q5_K_M2.8 GB 4.5 GB
Fits Get file →
Q6_K3.2 GB 4.8 GB
Fits Get file →
Q8_04.1 GB 5.8 GB
Fits Get file →
F167.8 GB 9.4 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (file size 1.73 GB, estimated 3.4 GB memory at 8k context). For 12 GB VRAM, Q4_K_M (2.49 GB, 4.1 GB memory) is a good balance. For 24 GB VRAM, Q8_0 (4.13 GB, 5.8 GB memory) offers higher precision.

What GPU do you need for Sarvam Translate?

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_S (2.4 GB file, about 4.0 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (4.1 GB file, about 5.8 GB in use).
  • 8 GB VRAM: the best quant that fits is Q8_0 (4.1 GB file, about 5.8 GB in use).
  • 12 GB VRAM: the best quant that fits is F16 (7.8 GB file, about 9.4 GB in use).

Which Sarvam Translate quant should you choose?

Q4_K_M is the usual starting point for Sarvam Translate: it balances file size and quality well. Here it is a 2.5 GB file.

The Q8_0 file is 4.1 GB, about 1.7 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 1.7 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

Architecturegemma3
Parameters3.88B
Layers34
Trained context131,072 tokens
KV cache (fp16)136 KB per token

Source

Quantized bymradermacher
FamilySarvam
Licensegpl-3.0
Downloads (HF)942
Last updated2026-10-13

Frequently asked questions about Sarvam Translate

How much VRAM does Sarvam Translate need?

About 4.1 GB at Q4_K_M with an 8K context window. That is the 2.5 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 Sarvam Translate on an 8 GB GPU?

Yes, with Q8_0: the 4.1 GB file needs about 5.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Sarvam Translate GGUF?

It depends on the quant: Q2_K is 1.7 GB, Q3_K_S is 1.9 GB, Q3_K_M is 2.1 GB, Q3_K_L is 2.2 GB, IQ4_XS is 2.3 GB, Q4_K_S is 2.4 GB and 6 more.

What context length does Sarvam Translate support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.6 GB at 32K.

Where can I download Sarvam Translate GGUF?

The files are on Hugging Face in the mradermacher/sarvam-translate-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 Sarvam Translate use?

Hugging Face lists the license as gpl-3.0. 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 Sarvam Translate on 8 GB VRAM?

Yes, the Q2_K quant uses 1.73 GB file size and requires about 3.4 GB VRAM at 8k context, fitting within 8 GB.

Does Sarvam Translate support Indic‑to‑Indic translation?

The open‑source checkpoint only supports English to Indian languages and vice‑versa; Indic‑to‑Indic translation requires a pivot through English.

Is there vision support?

The model has detected vision abilities but does not provide native vision processing; it focuses on text translation.

Other GGUFs of sarvam-translate

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Sarvam models

All Sarvam →

Other small models

All small models →