text model · MiniMax · Android
Can I run MiniMax M3 on Samsung Galaxy S25 Ultra (16GB, 1TB config only)?
No. MiniMax M3 needs ~269.8 GB even at Q4_K_M, but Samsung Galaxy S25 Ultra (16GB, 1TB config only) only has ~12 GB usable.
Needs ~269.8 GB even at Q4_K_M, but only ~12 GB is usable.
That figure is at a 4k context and moves about ±15% as context length changes.
The gap is about 257.8 GB: MiniMax M3 needs roughly 269.8 GB at Q4_K_M and Samsung Galaxy S25 Ultra (16GB, 1TB config only) leaves only about 12 GB usable for a model. No single tracked device has enough memory; MiniMax M3 needs a multi-GPU or high-memory rig.
- Q4_K_M needed
- ~269.8 GB
- Usable on device
- ~12 GB
- Device memory
- 16 GB
Which quant fits
How to run it
On Android use PocketPal AI (Polished app, download GGUF and run offline.).
- Parameters
- 428B (MoE, 23B active)
- Q4_K_M size
- 264.03 GB
- Q8_0 size
- 452.71 GB
- Context
- 1000k
- Memory
- 16 GB ram
- Usable for weights
- ~12 GB
- Power draw
- ~8 W
- Best runtime
- llama.cpp (PocketPal) or MLC-LLM (Adreno GPU path)
What you can run instead
FAQ
Can Samsung Galaxy S25 Ultra (16GB, 1TB config only) run MiniMax M3?
No. MiniMax M3 needs ~269.8 GB even at Q4_K_M, but Samsung Galaxy S25 Ultra (16GB, 1TB config only) only has ~12 GB usable.
How much memory does MiniMax M3 need?
Samsung Galaxy S25 Ultra (16GB, 1TB config only) does not have enough memory. At Q4_K_M the weights are ~264.03 GB; with KV cache and runtime overhead, budget ~269.8 GB at a 4k context. It is a Mixture-of-Experts model (428B total / 23B active), so all experts must stay in memory; memory tracks total params, not active params.
What is the best tool to run MiniMax M3 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. [](https://localmodel.run/can-i-run/minimax-m3/samsung-s25-ultra-16gb) Sources
- gadgetversus.com
- github.com/a-ghorbani
- github.com/google-ai-edge
- github.com/mlc-ai
- github.com/shubham0204
- github.com/Vali-98
- gsmarena.com/samsung_galaxy_s25_ultra-12750.php
- gsmarena.com/samsung_galaxy_s25_ultra-13322.php
- huggingface.co/MiniMaxAI
- huggingface.co/unsloth
- layla-network.ai
- sammyfans.com
- sammyguru.com
- samsung.com
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.