EM

Source Hf Embeddinggemma 2

EmbeddinggemmaGGUF by ngquocvinhUpdated 2026-10-08
Download on Hugging Face
Parameters271M
Context256K
Layers24
Architecturegemma-embedding2
Smallest quant0.1 GB
Quick answer

Source Hf Embeddinggemma 2 needs about 2.3 GB of memory at Q4_K_M (a 0.2 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 26 quants, from 0.1 GB (Q1_0) to 0.3 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.

EmbeddingGemma 2 – multimodal embedding model with 768‑d vectors

EmbeddingGemma 2 maps text, code, images, video, and audio into a unified 768‑dimensional space. It supports an 8,192‑token context window and offers modular vision and audio encoders. The model is open source under Apache 2.0.

Good at

  • Multimodal embeddings for search and retrieval
  • Low‑latency on consumer hardware
  • Supports 100+ languages and code embeddings

Limits

  • Requires 8K context window for full capability
  • Vision and audio encoders are optional but increase memory
  • No native vision support in the text‑only quant files

Best for

  • Embedding‑based search engines
  • Retrieval‑augmented generation pipelines
  • Multilingual semantic similarity tasks

Who it’s for. Developers building on‑device search or RAG applications that need fast, multilingual embeddings. The model is suitable for mobile and laptop deployments where low latency is critical.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q1_00.1 GB 2.2 GB
Fits Get file →
IQ1_S0.1 GB 2.2 GB
Fits Get file →
IQ1_M0.1 GB 2.2 GB
Fits Get file →
IQ2_XXS0.1 GB 2.2 GB
Fits Get file →
IQ2_XS0.1 GB 2.2 GB
Fits Get file →
Q2_K_S0.1 GB 2.2 GB
Fits Get file →
Q2_K0.1 GB 2.2 GB
Fits Get file →
IQ2_S0.1 GB 2.2 GB
Fits Get file →
IQ2_M0.1 GB 2.2 GB
Fits Get file →
TQ1_00.1 GB 2.2 GB
Fits Get file →
IQ3_XXS0.1 GB 2.2 GB
Fits Get file →
TQ2_00.1 GB 2.2 GB
Fits Get file →
IQ3_XS0.1 GB 2.3 GB
Fits Get file →
IQ3_S0.1 GB 2.3 GB
Fits Get file →
Q3_K_S0.1 GB 2.3 GB
Fits Get file →
IQ3_M0.1 GB 2.3 GB
Fits Get file →
Q3_K_M0.1 GB 2.3 GB
Fits Get file →
Q3_K_L0.2 GB 2.3 GB
Fits Get file →
IQ4_XS0.2 GB 2.3 GB
Fits Get file →
IQ4_NL0.2 GB 2.3 GB
Fits Get file →
Q4_K_S0.2 GB 2.3 GB
Fits Get file →
Q4_K_M0.2 GB 2.3 GB
Fits Get file →
Q5_K_S0.2 GB 2.3 GB
Fits Get file →
Q5_K_M0.2 GB 2.3 GB
Fits Get file →
Q6_K0.2 GB 2.4 GB
Fits Get file →
Q8_00.3 GB 2.4 GB
Fits Get file →

For 8 GB VRAM, use Q4_K_M or Q5_K_S (≈0.18–0.21 GB). For 4 GB VRAM, Q3_K_S or IQ3_S (≈0.14 GB) provides a good trade‑off. Larger models like Q6_K or Q8_0 require more than 8 GB VRAM and are not recommended for low‑memory devices.

What GPU do you need for Source Hf Embeddinggemma 2?

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 Q8_0 (0.3 GB file, about 2.4 GB in use).

Which Source Hf Embeddinggemma 2 quant should you choose?

Q4_K_M is the usual starting point for Source Hf Embeddinggemma 2: it balances file size and quality well. Here it is a 0.2 GB file.

The Q8_0 file is 0.3 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 Q1_0 at 0.1 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturegemma-embedding2
Parameters271M
Layers24
Trained context262,144 tokens
KV cache (fp16)192 KB per token

Source

Quantized byngquocvinh
Licenseapache-2.0
Downloads (HF)7,768
Last updated2026-10-08

Frequently asked questions about Source Hf Embeddinggemma 2

How much VRAM does Source Hf Embeddinggemma 2 need?

About 2.3 GB at Q4_K_M with an 8K context window. That is the 0.2 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 Source Hf Embeddinggemma 2 on an 8 GB GPU?

Yes, with Q8_0: the 0.3 GB file needs about 2.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Source Hf Embeddinggemma 2 GGUF?

It depends on the quant: Q1_0 is 0.1 GB, IQ1_S is 0.1 GB, IQ1_M is 0.1 GB, IQ2_XXS is 0.1 GB, IQ2_XS is 0.1 GB, Q2_K_S is 0.1 GB and 20 more.

What context length does Source Hf Embeddinggemma 2 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.4 GB at 32K.

Where can I download Source Hf Embeddinggemma 2 GGUF?

The files are on Hugging Face in the ngquocvinh/EmbeddingGemma-2-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 Source Hf Embeddinggemma 2 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 EmbeddingGemma 2 on 4 GB VRAM?

Yes, using Q3_K_S or IQ3_S quantizations, which are about 0.14 GB in size and require roughly 2.3 GB of VRAM at 8K context.

Does the model support image inputs?

The quantized text backbone does not include vision; a separate vision encoder is optional and must be loaded separately.

What license applies to this model?

The model is released under the Apache 2.0 license, allowing free use and modification.

Other GGUFs of embeddinggemma-2

Similar models

Browse all →

More from ngquocvinh

All by ngquocvinh →

More Embeddinggemma models

All Embeddinggemma →

Other tiny models

All tiny models →