text model · Gemma · Android

Can I run Gemma 4 12B on Samsung Galaxy S25 Ultra (16GB, 1TB config only)?

Compatibility verdict VRAM threshold engine

Yes, it runs slow on this hardware ~6 tok/s est.

Yes. Gemma 4 12B runs on Samsung Galaxy S25 Ultra (16GB, 1TB config only) at Q4_K_M (~8.7 GB of ~12 GB usable).

Needs ~8.7 GB Device usable ~12 GB

Runs at Q4_K_M using ~8.7 GB of ~12 GB usable.

That figure is at a 4k context and moves about ±15% as context length changes. Samsung Galaxy S25 Ultra (16GB, 1TB config only) leaves ~3.3 GB of headroom.

Q4_K_M needed: ~8.7 GB
Usable on device: ~12 GB
Device memory: 16 GB
Best quant: Q4_K_M

Share on X Share on Reddit

Which quant fits

Quant ladder vs ~12 GB usable

Q2_K

~6.6 GB

Q3_K_M

~7.5 GB

Q4_K_M

~8.7 GB

Q5_K_M

~10.2 GB

Q6_K

~11.4 GB

Q8_0

~14.3 GB

FP16

~25.4 GB

The line marks Samsung Galaxy S25 Ultra (16GB, 1TB config only)'s ~12 GB budget; rungs past it are too large.

Running cost · estimate

Power & economics est.

Power draw: ~8 W
Electricity / 1M tokens: ~$0.06
Pays for itself after: ~3,680M tok

At ~$0.15/kWh and the estimated ~6 tok/s, a million generated tokens costs about $0.06 in electricity. Versus a hosted API at ~$0.5 per million tokens, the ~$1,619 Samsung Galaxy S25 Ultra (16GB, 1TB config only) pays for itself after roughly 3,680 million tokens, so local hardware is mostly a fixed cost, not a per-token one. TDP is the peak draw, so this is an upper bound. Assumptions.

How to run it

On Android use PocketPal AI (Polished app, download GGUF and run offline.).

Model Gemma

Parameters: 12B
Q4_K_M size: 7.12 GB
Q8_0 size: 12.67 GB
Context: 256k
Ollama tag: gemma4:12b

Full Gemma 4 12B requirements →

Device Android

Memory: 16 GB ram
Usable for weights: ~12 GB
Power draw: ~8 W
Best runtime: llama.cpp (PocketPal) or MLC-LLM (Adreno GPU path)

Best models for Samsung Galaxy S25 Ultra (16GB, 1TB config only) →

You could also run

Run Gemma 4 12B on other hardware

Nvidia GeForce RTX 3060 (12GB)Nvidia GeForce RTX 4070 (12GB)Apple M2 (16GB)Apple M4 (16GB)Nvidia GeForce RTX 4060 Ti (16GB)Nvidia GeForce RTX 4080 (16GB)

FAQ

Can Samsung Galaxy S25 Ultra (16GB, 1TB config only) run Gemma 4 12B?

Yes. Gemma 4 12B runs on Samsung Galaxy S25 Ultra (16GB, 1TB config only) at Q4_K_M (~8.7 GB of ~12 GB usable).

How much memory does Gemma 4 12B need?

Samsung Galaxy S25 Ultra (16GB, 1TB config only) has room to spare. At Q4_K_M the weights are ~7.12 GB; with KV cache and runtime overhead, budget ~8.7 GB at a 4k context.

What is the best tool to run Gemma 4 12B on Android?

On Android, PocketPal AI (Polished app, download GGUF and run offline.) is the go-to option. NPU acceleration is limited and chip-specific; most apps run on CPU. Expect 1B-4B class.

Embed this

A live badge for your model card or README, updated as the data is.

Markdown

[![Gemma 4 12B on Samsung Galaxy S25 Ultra (16GB, 1TB config only)](https://localmodel.run/badge/gemma-4-12b/samsung-s25-ultra-16gb.svg)](https://localmodel.run/can-i-run/gemma-4-12b/samsung-s25-ultra-16gb)

Sources

Weights are measured from GGUF files; KV cache and overhead are computed, so totals can vary ~15% with context and runtime. Any tok/s is a bandwidth estimate. See methodology.