SAVRN
Search Contact SAVRN

Open-weight model · Any to any

gemma-4-E4B-it-qat-w4a16-ct

by Google google/gemma-4-E4B-it-qat-w4a16-ct

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output.

Parameters8.7B
Context131,072
Weights11.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads324.3k

Runs On

What it takes to serve gemma-4-E4B-it-qat-w4a16-ct (8.7B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 17.5 GB 21.0 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 8.7 GB 10.5 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 4.4 GB 5.2 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

SAVRN's Notes on gemma-4-E4B-it-qat-w4a16-ct

Five gigabytes, 5.2 to be exact, is what this Gemma 4 E4B build needs at 4-bit, and that number is why it exists. Google quantized it from google/gemma-4-E4B-it-qat-q4_0-unquantized. It takes text, image and audio input, writes text, and carries 8.7 billion parameters with a 131,072 token context. Even at 16-bit it needs 21 GB, so the cheapest rental we track, one MI300X at $1.85 an hour, leaves most of its 192 GB idle. Think small card, or several copies on one large card.

Apache 2.0 keeps the deployment simple: commercial use, modification and redistribution are allowed if you keep the license, copyright and NOTICE files and state significant changes, and it includes a patent grant. Check the context figure before planning around it: the family card cites up to 256K tokens, but this checkpoint's configuration reads 131,072 with a 512 token sliding window. No Index host prices it per token yet. The report is arXiv:2607.02770.

SAVRN Research, 2026-09-18

Model Card

By Google, published under apache-2.0, revision 6cd26aaa2357.

Hugging Face | GitHub | Launch Blog | Documentation | Technical Report
License: Apache 2.0 | Authors: Google DeepMind

Read the full model card (3,619 words)

Configuration

Architecture
Gemma4ForConditionalGeneration
Context length (tokens)
131,072
Layers
42
Hidden size
2,560
Feed-forward size
10,240
Attention heads
8
Key/value heads
2
Head dimension
256
Vocabulary size
262,144
Sliding window (tokens)
512
Model type
gemma4
Quantization
compressed-tensors

Identity and Version

Repository
google/gemma-4-E4B-it-qat-w4a16-ct
Publisher
Google
Task
Any to any
Modality
Multimodal
Library
transformers
Parameters
8.7B parameters
Languages
Not stated by the source
Revision
6cd26aaa2357fb2bad8c51699a7558a4d1a965bb
First published
2026-06-04
Last updated
2026-07-20

Files and Weights

9 files, 11.5 GB in total. The weights are 1 file totalling 11.5 GB in safetensors.

Weights1 file · 11.5 GB
Configuration3 files · 24.3 KB
Tokenizer2 files · 32.2 MB
Documentation1 file · 29.4 KB
Other1 file · 18.6 KB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model.safetensorsWeights11.5 GB 19d89d2a4b3e
config.jsonConfiguration22.4 KB —
generation_config.jsonConfiguration208 B —
processor_config.jsonConfiguration1.7 KB —
README.mdDocumentation29.4 KB —
chat_template.jinjaOther18.6 KB —
.gitattributesRepository1.6 KB —
tokenizer.jsonTokenizer32.2 MB cc8d3a0ce364
tokenizer_config.jsonTokenizer3.7 KB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
11.5 GB
Download from Google

Released by Google through Kaggle. Read the license.

Built From

  • Derived from google/gemma-4-E4B-it-qat-q4_0-unquantized
  • Described by arXiv:2607.02770
  • Quantized from google/gemma-4-E4B-it-qat-q4_0-unquantized

Memory Requirements

PrecisionWeights in memory
As published11.5 GB
16-bit17.5 GB
8-bit8.7 GB
4-bit4.4 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Compare gemma-4-E4B-it-qat-w4a16-ct

Questions About gemma-4-E4B-it-qat-w4a16-ct

How much GPU memory does gemma-4-E4B-it-qat-w4a16-ct need?

About 21 GB at 16-bit and 5.2 GB at 4-bit: the weights (8.7B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run gemma-4-E4B-it-qat-w4a16-ct on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use gemma-4-E4B-it-qat-w4a16-ct commercially?

Yes. gemma-4-E4B-it-qat-w4a16-ct is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

What is gemma-4-E4B-it-qat-w4a16-ct's context length?

131,072 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Any to any

MiniCPM-o-2_6

OpenBMB

[2025.06.20] Our official ollama repository is released. Try our latest models with one click！ [2025.03.01] RLAIF-V, which is the alignment technique of MiniCPM-o, is accepted by CVPR 2025！The code, dataset, paper are open-sourced! [2025.01.24] MiniCPM-o 2.6 technical report is released! See Here. [2025.01.19] MiniCPM-o tops GitHub Trending and reaches top-2 on Hugging Face Trending! MiniCPM-o 2.6 is the latest and most capable model in the MiniCPM-o series. The model is built in an end-to-end fashion based on SigLip-400M, Whisper-medium-300M, ChatTTS-200M, and Qwen2.5-7B with a total of 8B parameters. It exhibits a significant performance improvement over MiniCPM-V 2.6, and introduces new…

Open weights apache-2.0 8.7B parameters 32,768 tokens transformers
View model

Model · Any to any

MiniCPM-o-4_5

OpenBMB

A Gemini 2.5 Flash Level MLLM for Vision, Speech, and Full-Duplex Mulitmodal Live Streaming on | CaseBook(Audio, Omni Full-Duplex) MiniCPM-o 4.5 is the latest and most capable model in the MiniCPM-o series. The model is built in an end-to-end fashion based on SigLip2, Whisper-medium, CosyVoice2, and Qwen3-8B with a total of 9B parameters. It exhibits a significant performance improvement, and introduces new features for full-duplex multimodal live streaming. Notable features of MiniCPM-o 4.5 include: - Leading Visual Capability. MiniCPM-o 4.5 achieves an average score of 77.6 on OpenCompass, a comprehensive evaluation of 8 popular benchmarks. With only 9B parameters, it surpasses widely…

Open weights apache-2.0 9.4B parameters 40,960 tokens transformers
View model

Model · Any to any

MiniCPM-o-4_5-awq

OpenBMB

A Gemini 2.5 Flash Level MLLM for Vision, Speech, and Full-Duplex Mulitmodal Live Streaming on | CaseBook(Audio, Omni Full-Duplex) MiniCPM-o 4.5 is the latest and most capable model in the MiniCPM-o series. The model is built in an end-to-end fashion based on SigLip2, Whisper-medium, CosyVoice2, and Qwen3-8B with a total of 9B parameters. It exhibits a significant performance improvement, and introduces new features for full-duplex multimodal live streaming. Notable features of MiniCPM-o 4.5 include: - Leading Visual Capability. MiniCPM-o 4.5 achieves an average score of 77.6 on OpenCompass, a comprehensive evaluation of 8 popular benchmarks. With only 9B parameters, it surpasses widely…

Open weights apache-2.0 9.4B parameters 40,960 tokens transformers
View model

Model · Any to any

gemma-4-E4B-it

Google

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: E2B, E4B, 12B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from…

Open weights apache-2.0 8B parameters 131,072 tokens transformers
View model

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 4-bit quantized version of gemma-4-E4B-it using MLX, optimized for Apple Silicon. Special thanks to the Apple Machine Learning Research team for creating MLX. LM Studio is not the creator, originator, or owner of any Model featured in the Community Model Program. Each Community Model is created and provided by third parties. LM Studio does not endorse, support, represent or guarantee the completeness, truthfulness, accuracy, or reliability of any Community Model. You understand that Community Models can produce content that might be offensive, harmful…

Open weights apache-2.0 8B parameters 131,072 tokens transformers
View model

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 8-bit quantized version of gemma-4-E4B-it using MLX, optimized for Apple Silicon. Special thanks to the Apple Machine Learning Research team for creating MLX. LM Studio is not the creator, originator, or owner of any Model featured in the Community Model Program. Each Community Model is created and provided by third parties. LM Studio does not endorse, support, represent or guarantee the completeness, truthfulness, accuracy, or reliability of any Community Model. You understand that Community Models can produce content that might be offensive, harmful…

Open weights apache-2.0 8B parameters 131,072 tokens transformers
View model