text model · Qwen3 · Android

Can I run Qwen3 14B on Samsung Galaxy S25 Ultra (16GB, 1TB config only)?

Compatibility verdict VRAM threshold engine

Yes, it runs slow on this hardware ~5 tok/s est.

Yes. Qwen3 14B runs on Samsung Galaxy S25 Ultra (16GB, 1TB config only) at Q4_K_M (~10.7 GB of ~12 GB usable).

Needs ~10.7 GB Device usable ~12 GB

Runs at Q4_K_M using ~10.7 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 ~1.3 GB of headroom.

Q4_K_M needed: ~10.7 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

~7.6 GB

Q3_K_M

~8.5 GB

Q4_K_M

~10.7 GB

Q5_K_M

~11.7 GB

Q6_K

~13.2 GB

Q8_0

~17.4 GB

FP16

~31.7 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: ~$0.07
Pays for itself after: ~3,765M tok

At ~$0.15/kWh and the estimated ~5 tok/s, a million generated tokens costs about $0.07 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 3,765 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 Qwen3

Parameters: 14B
Q4_K_M size: 9 GB
Q8_0 size: 15.7 GB
Context: 32k
Ollama tag: qwen3:14b

Full Qwen3 14B 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 Qwen3 14B on other hardware

Nvidia GeForce RTX 3060 (12GB)Nvidia GeForce RTX 4070 (12GB)Nvidia GeForce RTX 4060 Ti (16GB)Nvidia GeForce RTX 4080 (16GB)16GB RAM Laptop (CPU/iGPU only)iPad Pro M4 (16GB, 1TB/2TB config)

FAQ

Can Samsung Galaxy S25 Ultra (16GB, 1TB config only) run Qwen3 14B?

Yes. Qwen3 14B runs on Samsung Galaxy S25 Ultra (16GB, 1TB config only) at Q4_K_M (~10.7 GB of ~12 GB usable).

How much memory does Qwen3 14B need?

Samsung Galaxy S25 Ultra (16GB, 1TB config only) has room to spare. At Q4_K_M the weights are ~9 GB; with KV cache and runtime overhead, budget ~10.7 GB at a 4k context.

What is the best tool to run Qwen3 14B 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

[![Qwen3 14B on Samsung Galaxy S25 Ultra (16GB, 1TB config only)](https://localmodel.run/badge/qwen3-14b/samsung-s25-ultra-16gb.svg)](https://localmodel.run/can-i-run/qwen3-14b/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.