# gemma-4-12B-it by Google: Open-Weight Model
Source: https://savrn.com/models/gemma-4-12b-it
Markdown alternate of the page above; the site index is https://savrn.com/llms.txt

---

## Runs On

What it takes to serve gemma-4-12B-it (12B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
| --- | --- | --- | --- | --- | --- |
| 16-bit | 23.9 GB | 28.7 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 |
| 8-bit | 12.0 GB | 14.4 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 |
| 4-bit | 6.0 GB | 7.2 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/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](https://savrn.com/ai-index/pricing/gpus), read Oct 24, 2026.

[gemma-4-12B-it on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/gemma-4-12b-it/gpus)

## SAVRN's Notes on gemma-4-12B-it

Text and images in, audio too on this size, text out, 140-plus languages, a 256K-token window: that is Google's instruction-tuned Gemma 4 12B. The footprint is 28.7 GB at 16-bit, 14.4 GB at 8-bit and 7.2 GB at 4-bit, all three on one MI300X at $1.85 per hour with 192 GB to work in, so the question is not which card but how much of it you leave for a 262,144-token context and its neighbors.

Nothing in Apache 2.0 stops you: commercial use, modification and redistribution, with license and NOTICE files kept and changes stated. It derives from google/gemma-4-12B, the pre-trained base, so start there for your own tuning, and read arXiv:2607.02770 first. The scores here, 78.8 on GPQA diamond among them, are model card numbers, not ours; run your own documents. No Index host prices it by the token; the comparison is the card hour.

SAVRN Research, 2026-09-18

## Model Card

By Google, published under apache-2.0, revision 707f0a3b8a3c.

[Hugging Face](https://huggingface.co/collections/google/gemma-4) | [GitHub](https://github.com/google-gemma) | [Launch Blog](https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/) | [Documentation](https://ai.google.dev/gemma/docs/core) | [Technical Report](https://savrn.com/papers/gemma-4-technical-report) License: [Apache 2.0](https://ai.google.dev/gemma/docs/gemma_4_license) | Authors: [Google DeepMind](https://deepmind.google/models/gemma/)

[!Note] This model card is for the Gemma 4 12B Unified model, which is part of the Gemma 4 family of open models. Built with the same multimodal functionality as Gemma 4 E2B and E4B (text, audio, image, and video inputs), it brings native audio and vision understanding directly to local environments without the need for separate encoders. This unified approach to multimodality makes the model encoder-free, offering a deployment size that is perfect for consumer devices and streamlined local execution.

[Read the full model card (3,506 words)](https://savrn.com/models/gemma-4-12b-it/card)

## Configuration

Architecture

Gemma4UnifiedForConditionalGeneration

Context length (tokens)

262,144

Layers

48

Hidden size

3,840

Feed-forward size

15,360

Attention heads

16

Key/value heads

8

Head dimension

256

Vocabulary size

262,144

Sliding window (tokens)

1,024

Model type

gemma4_unified

## Identity and Version

Repository

google/gemma-4-12B-it

Publisher

Google

Task

Any to any

Modality

Multimodal

Library

transformers

Parameters

12B parameters

Languages

Not stated by the source

Revision

707f0a3b8a3c7ad586ed01e27eafbad8a27dd0f7

First published

2026-05-23

Last updated

2026-07-20

## Files and Weights

9 files, 24.0 GB in total. The weights are 1 file totalling 23.9 GB in safetensors.

Weights1 file · 23.9 GB

Configuration3 files · 6.1 KB

Tokenizer2 files · 32.2 MB

Documentation1 file · 28.5 KB

Other1 file · 18.7 KB

Repository1 file · 1.6 KB

Every file

| File | Type | Size | SHA-256 |
| --- | --- | --- | --- |
| model.safetensors | Weights | 23.9 GB | 5a84cb313260 |
| config.json | Configuration | 4.4 KB | — |
| generation_config.json | Configuration | 260 B | — |
| processor_config.json | Configuration | 1.4 KB | — |
| README.md | Documentation | 28.5 KB | — |
| chat_template.jinja | Other | 18.7 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 32.2 MB | cc8d3a0ce364 |
| tokenizer_config.json | Tokenizer | 3.1 KB | — |

## License and Download

License

apache-2.0

Access

Open weights, no gate

Download size

23.9 GB

[Download from Google](https://www.kaggle.com/models/google/gemma-4)

Released by Google through Kaggle. [Read the license](https://www.apache.org/licenses/LICENSE-2.0).

## Built From

- Derived from [google/gemma-4-12B](https://savrn.com/models/gemma-4-12b)
- Described by [arXiv:2607.02770](https://savrn.com/papers/gemma-4-technical-report)

## Evaluations

Each result is shown as reported, with the conditions its reporter stated. None is a SAVRN measurement. A comparison lines two results up only when their configuration, unit and setup are all stated and identical.

| Benchmark | Conditions | Result | Reported by | Revision | Date |
| --- | --- | --- | --- | --- | --- |
| Idavidrein/gpqa | Task diamondMetric diamondComparison conditions not established | 78.8 | [Model Card](https://huggingface.co/google/gemma-4-12B-it) Reported by a third party | Evaluated revision not stated | 2026-05-23 |
| MMMU/MMMU_Pro | Task mmmu_proMetric mmmu_proComparison conditions not established | 69.1 | [Model Card](https://huggingface.co/google/gemma-4-12B-it) Reported by a third party | Evaluated revision not stated | 2026-05-23 |
| MathArena/aime_2026 | Task MathArena/aime_2026Metric MathArena/aime_2026Setup No toolsComparison conditions not established | 77.5 | [Model Card](https://huggingface.co/google/gemma-4-12B-it) Reported by a third party | Evaluated revision not stated | 2026-05-23 |
| TIGER-Lab/MMLU-Pro | Task mmlu_proMetric mmlu_proComparison conditions not established | 77.2 | [Model Card](https://huggingface.co/google/gemma-4-12B-it) Reported by a third party | Evaluated revision not stated | 2026-05-23 |
| cais/hle | Task hleMetric hleSetup No toolsComparison conditions not established | 5.2 | [Model Card](https://huggingface.co/google/gemma-4-12B-it) Reported by a third party | Evaluated revision not stated | 2026-05-23 |
| llamaindex/ExtractBench | Task longMetric longSetup Pipeline name: gemma4_12b_vllm_extract_oneshot_structured_output_fileComparison conditions not established | 13.16 | [ExtractBench](https://huggingface.co/datasets/llamaindex/ExtractBench) Reported by a third party | Evaluated revision not stated | 2026-08-24 |
| llamaindex/ExtractBench | Task meanMetric meanSetup Pipeline name: gemma4_12b_vllm_extract_oneshot_structured_output_fileComparison conditions not established | 45.39 | [ExtractBench](https://huggingface.co/datasets/llamaindex/ExtractBench) Reported by a third party | Evaluated revision not stated | 2026-08-24 |
| llamaindex/ExtractBench | Task mediumMetric mediumSetup Pipeline name: gemma4_12b_vllm_extract_oneshot_structured_output_fileComparison conditions not established | 30.82 | [ExtractBench](https://huggingface.co/datasets/llamaindex/ExtractBench) Reported by a third party | Evaluated revision not stated | 2026-08-24 |
| llamaindex/ExtractBench | Task shortMetric shortSetup Pipeline name: gemma4_12b_vllm_extract_oneshot_structured_output_fileComparison conditions not established | 53.62 | [ExtractBench](https://huggingface.co/datasets/llamaindex/ExtractBench) Reported by a third party | Evaluated revision not stated | 2026-08-24 |

## Memory Requirements

| Precision | Weights in memory |
| --- | --- |
| As published | 23.9 GB |
| 16-bit | 23.9 GB |
| 8-bit | 12.0 GB |
| 4-bit | 6.0 GB |

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

## Built on This Model

- Quantized from[gemma-4-12b-it-GGUF](https://savrn.com/models/gemma-4-12b-it-gguf)
- Derived from[gemma-4-12b-it-GGUF](https://savrn.com/models/gemma-4-12b-it-gguf)
- Quantized from[Gemma4-12B-QAT-Uncensored-HauhauCS-Balanced](https://savrn.com/models/gemma4-12b-qat-uncensored-hauhaucs-balanced)
- Derived from[Gemma4-12B-QAT-Uncensored-HauhauCS-Balanced](https://savrn.com/models/gemma4-12b-qat-uncensored-hauhaucs-balanced)
- Quantized from[gemma-4-12B-it-FP8-Dynamic](https://savrn.com/models/gemma-4-12b-it-fp8-dynamic)
- Derived from[gemma-4-12B-it-FP8-Dynamic](https://savrn.com/models/gemma-4-12b-it-fp8-dynamic)
- Quantized from[gemma-4-12B-it-MLX-4bit](https://savrn.com/models/gemma-4-12b-it-mlx-4bit)
- Derived from[gemma-4-12B-it-MLX-4bit](https://savrn.com/models/gemma-4-12b-it-mlx-4bit)
- Quantized from[gemma-4-12B-it-MLX-8bit](https://savrn.com/models/gemma-4-12b-it-mlx-8bit)
- Derived from[gemma-4-12B-it-MLX-8bit](https://savrn.com/models/gemma-4-12b-it-mlx-8bit)
- Quantized from[gemma-4-12B-it-MLX-6bit](https://savrn.com/models/gemma-4-12b-it-mlx-6bit)
- Derived from[gemma-4-12B-it-MLX-6bit](https://savrn.com/models/gemma-4-12b-it-mlx-6bit)
- Quantized from[gemma-4-12B-it-MLX-5bit](https://savrn.com/models/gemma-4-12b-it-mlx-5bit)
- Derived from[gemma-4-12B-it-MLX-5bit](https://savrn.com/models/gemma-4-12b-it-mlx-5bit)
- Quantized from[gemma-4-12B-it-AWQ-INT4](https://savrn.com/models/gemma-4-12b-it-awq-int4)
- Derived from[gemma-4-12B-it-AWQ-INT4](https://savrn.com/models/gemma-4-12b-it-awq-int4)
- Quantized from[gemma-4-12B-it-FP8-dynamic](https://savrn.com/models/gemma-4-12b-it-fp8-dynamic-2)
- Derived from[gemma-4-12B-it-FP8-dynamic](https://savrn.com/models/gemma-4-12b-it-fp8-dynamic-2)
- Quantized from[gemma-4-12B-it-GGUF](https://savrn.com/models/gemma-4-12b-it-gguf-2)
- Derived from[gemma-4-12B-it-GGUF](https://savrn.com/models/gemma-4-12b-it-gguf-2)
- Quantized from[gemma-gguf](https://savrn.com/models/gemma-gguf)
- Derived from[gemma-gguf](https://savrn.com/models/gemma-gguf)
- Quantized from[gemma-4-12b-it-coreml-128k](https://savrn.com/models/gemma-4-12b-it-coreml-128k)
- Derived from[gemma-4-12b-it-coreml-128k](https://savrn.com/models/gemma-4-12b-it-coreml-128k)
- Adapter of[gemma-4-12b-roleplay-sft-regmix-v2](https://savrn.com/models/gemma-4-12b-roleplay-sft-regmix-v2)
- Derived from[gemma-4-12b-roleplay-sft-regmix-v2](https://savrn.com/models/gemma-4-12b-roleplay-sft-regmix-v2)
- Quantized from[gemma-4-12B-it-Pollard](https://savrn.com/models/gemma-4-12b-it-pollard)
- Derived from[gemma-4-12B-it-Pollard](https://savrn.com/models/gemma-4-12b-it-pollard)
- Derived from[Jev-Omni](https://savrn.com/models/jev-omni)
- Derived from[SDRPN-Gemma-4-12B](https://savrn.com/models/sdrpn-gemma-4-12b)
- Adapter of[gemma-4-12b-pixelproof-lora-adapters](https://savrn.com/models/gemma-4-12b-pixelproof-lora-adapters)
- Derived from[gemma-4-12b-pixelproof-lora-adapters](https://savrn.com/models/gemma-4-12b-pixelproof-lora-adapters)
- Derived from[Clemma-12B](https://savrn.com/models/clemma-12b)
- Derived from[Huihui-gemma-4-12B-it-abliterated](https://savrn.com/models/huihui-gemma-4-12b-it-abliterated)
- Derived from[gemma4-12b-biblical-qlora](https://savrn.com/models/gemma4-12b-biblical-qlora)
- Quantized from[gemma-4-12B-it-JNS](https://savrn.com/models/gemma-4-12b-it-jns)
- Derived from[gemma-4-12B-it-JNS](https://savrn.com/models/gemma-4-12b-it-jns)
- Derived from[gemma4-12b-biblical-qlora-gt](https://savrn.com/models/gemma4-12b-biblical-qlora-gt)
- Derived from[Huihui-gemma-4-12B-it-abliterated](https://savrn.com/models/huihui-gemma-4-12b-it-abliterated-2)
- Adapter of[rlport-sft_jordgm_own](https://savrn.com/models/rlport-sft-jordgm-own)

## Compare gemma-4-12B-it

- [gemma-4-12B-it vs gemma-4-E4B-it-qat-w4a16-ct](https://savrn.com/models/compare/gemma-4-12b-it-vs-gemma-4-e4b-it-qat-w4a16-ct)
- [gemma-4-12B-it vs MiniCPM-o-4_5](https://savrn.com/models/compare/gemma-4-12b-it-vs-minicpm-o-4-5)
- [gemma-4-12B-it vs Qwen2.5-Omni-7B](https://savrn.com/models/compare/gemma-4-12b-it-vs-qwen2-5-omni-7b)

## Questions About gemma-4-12B-it

### How much GPU memory does gemma-4-12B-it need?

About 28.7 GB at 16-bit and 7.2 GB at 4-bit: the weights (12B parameters) plus a working margin. A long context needs more.

### What is the cheapest GPU to run gemma-4-12B-it 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-4-12B-it commercially?

Yes. gemma-4-12B-it 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-4-12B-it's context length?

262,144 tokens, from the maximum position embeddings in its published configuration.

## Similar Models

Model · Any to any

### [gemma-4-12B-it-MLX-4bit](https://savrn.com/models/gemma-4-12b-it-mlx-4bit)

[LM Studio Community](https://savrn.com/model-publishers/lmstudio-community)

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 4-bit quantized version of gemma-4-12B-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 12B parameters 262,144 tokens transformers

[View model](https://savrn.com/models/gemma-4-12b-it-mlx-4bit)

Model · Any to any

### [gemma-4-12B-it-MLX-8bit](https://savrn.com/models/gemma-4-12b-it-mlx-8bit)

[LM Studio Community](https://savrn.com/model-publishers/lmstudio-community)

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 8-bit quantized version of gemma-4-12B-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 12B parameters 262,144 tokens transformers

[View model](https://savrn.com/models/gemma-4-12b-it-mlx-8bit)

Model · Any to any

### [gemma-4-12B-it-MLX-6bit](https://savrn.com/models/gemma-4-12b-it-mlx-6bit)

[LM Studio Community](https://savrn.com/model-publishers/lmstudio-community)

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 6-bit quantized version of gemma-4-12B-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 12B parameters 262,144 tokens transformers

[View model](https://savrn.com/models/gemma-4-12b-it-mlx-6bit)

Model · Any to any

### [gemma-4-12B-it-MLX-5bit](https://savrn.com/models/gemma-4-12b-it-mlx-5bit)

[LM Studio Community](https://savrn.com/model-publishers/lmstudio-community)

LM Studio Community models highlights program. Highlighting new & noteworthy models by the community. Join the conversation on Discord. 5-bit quantized version of gemma-4-12B-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 12B parameters 262,144 tokens transformers

[View model](https://savrn.com/models/gemma-4-12b-it-mlx-5bit)

Model · Any to any

### [gemma-4-12B](https://savrn.com/models/gemma-4-12b)

[Google](https://savrn.com/model-publishers/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 12B parameters 262,144 tokens transformers

[View model](https://savrn.com/models/gemma-4-12b)

Model · Any to any

### [Huihui-gemma-4-12B-it-abliterated](https://savrn.com/models/huihui-gemma-4-12b-it-abliterated)

[Isaac Jaxxon](https://savrn.com/model-publishers/timelesssoul)

This is an uncensored version of google/gemma-4-12B-it created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens. Note For this model, both the thinking mode and the non-thinking mode have been completely abliterated. Only layers 23-28 have been abliterated. Important Notice: The weights of the earliest released google/gemma-4-12B-it model have issues. We have already re-ablated and re-uploaded the model. If you have already downloaded it, please re-download. Please use the latest version of ollama You can use huihuiai/gemma-4-abliterated:12b…

Open weights apache-2.0 12B parameters 131,072 tokens transformers

[View model](https://savrn.com/models/huihui-gemma-4-12b-it-abliterated)

## Google

[All models and datasets](https://savrn.com/model-publishers/google)

## Versions

- [707f0a3b8a3c](https://savrn.com/models/gemma-4-12b-it/versions/707f0a3b8a3c) · current 2026-10-24

## Explore More

- [All any to any models](https://savrn.com/models/tasks/any-to-any)
- [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0)
- [Model comparisons](https://savrn.com/models/comparisons)
- [The model directory](https://savrn.com/models)
- [Open model prices by host](https://savrn.com/ai-index/pricing/open-models)

## Source

- Repository metadata, read 2026-10-19.
- [Google website](https://ai.google.dev/)
- [Hugging Face record](https://huggingface.co/google/gemma-4-12B-it)
- [How the hub is built](https://savrn.com/model-hub/methodology)