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 8 quants, from 0.1 GB (Q3_K_M) to 0.6 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.
Embeddinggemma 2: 740M‑parameter multimodal embedding model
Embeddinggemma 2 is a 740M‑parameter model that maps text, code, images, video, and audio into a unified 768‑dimensional vector space. It supports an 8K‑token context window and offers Matryoshka Representation Learning for truncated embeddings. The model is licensed under Apache 2.0.
Good at
- Multimodal embeddings for text, image, video, audio
- Low‑latency on consumer hardware
- Supports 100+ languages
Limits
- No vision support in quantized files
- Requires 8K context for full performance
- No explicit GPU acceleration info
Best for
- Semantic search on-device
- Retrieval‑augmented generation
- Multimodal similarity matching
Who it’s for. Developers building on‑device or edge applications that need fast, multilingual, multimodal embeddings. The model is suitable for search, RAG, classification, and clustering tasks where a unified embedding space is required.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_K_M | 0.1 GB | 2.3 GB | Fits | Get file → | |
| Q3_K_L | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q4_K_S | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q4_K_M | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q5_K_S | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q5_K_M | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q6_K | 0.2 GB | 2.4 GB | Fits | Get file → | |
| BF16 | 0.6 GB | 2.7 GB | Fits | Get file → |
For 8GB VRAM, use Q4_K_M or Q4_K_S (0.18 GB file, ~2.3 GB memory). For 4GB VRAM, Q3_K_M or Q3_K_L (0.15 GB file, ~2.3 GB memory). BF16 (0.56 GB file) requires ~2.7 GB memory and is best for higher precision needs.
What GPU do you need for 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 BF16 (0.6 GB file, about 2.7 GB in use).
Which Embeddinggemma 2 quant should you choose?
Q4_K_M is the usual starting point for Embeddinggemma 2: it balances file size and quality well. Here it is a 0.2 GB file.
If memory is tight, the smallest file is Q3_K_M at 0.1 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Embeddinggemma 2
How much VRAM does 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 Embeddinggemma 2 on an 8 GB GPU?
Yes, with BF16: the 0.6 GB file needs about 2.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Embeddinggemma 2 GGUF?
It depends on the quant: Q3_K_M is 0.1 GB, Q3_K_L is 0.2 GB, Q4_K_S is 0.2 GB, Q4_K_M is 0.2 GB, Q5_K_S is 0.2 GB, Q5_K_M is 0.2 GB and 2 more.
What context length does 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 Embeddinggemma 2 GGUF?
The files are on Hugging Face in the prithivMLmods/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 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 8GB VRAM?
Yes, using Q4_K_M or Q4_K_S quantizations which use about 2.3 GB of VRAM at 8K context.
Does the model support vision?
Yes, it includes a 170M vision encoder that can be loaded independently.
What context length does it support?
It supports an 8K‑token context window for full performance.