Can Llama 3.1 8B run on ROG Phone 9 Pro?
YES — Runs, barely
Grade A · 79/100 · formula estimate
What this means
The checked Q4_K_M version should run, but replies are likely to arrive slowly. The smaller Q3_K_M version is the safer choice.
Download the Q3_K_M version, which is about 4 GB. We expect it to use about 5.4 GB of the roughly 20 GB available to a model on this phone.
At the estimated speed, a roughly 300-word answer may take about 47 seconds to finish.
This result is calculated from the phone and model specifications. It has not been measured on this exact phone and setup.
7.1tokens/s
estimated · reading pace
6.3GB
needed at Q4_K_M
Q4_K_M
quant checked
A
grade · 79/100
needs 6.3 GBusable 20 GB
Snapdragon 8 Elite24 GB RAMCPU / GPUalso sold with 16 GB
Make it bearable
1
Drop to Q3_K_M (4 GB) — it runs comfortably here at ~8.6 tokens/s. Recommended over Q4_K_M.
2
Drop context to 2K — saves ~0.3 GB of KV cache and a little speed.
3
Close every other app before loading — the 13.7 GB headroom is real; Android may kill the app otherwise.
4
Expect throttling after ~10 min of sustained generation on a phone chassis.
See how fast it feels
Estimated at 7.1 tokens/s — reading pace. A ~300-word reply takes about 56 seconds on this phone.
Where the memory goes
| Component | Detail | GB |
|---|---|---|
| Model weights | Q4_K_M GGUF (4.9 GB) + mmap overhead | 5.1 |
| KV cache | 4K context window | 0.6 |
| Runtime | llama.cpp + app overhead | 0.6 |
| Total needed | at Q4_K_M, 4K context | 6.3 |
| Usable on ROG Phone 9 Pro | 24 GB RAM − Android system reserve | 20 |
| Headroom | left for the OS and your other apps | 13.7 |
Pick your quant
| Quant | Download | Verdict | Speed |
|---|---|---|---|
| Q2_K | 3.2 GB | ✓ Runs great | ~10.8 tokens/s |
| Q3_K_M BEST HERE | 4 GB | ✓ Runs great | ~8.6 tokens/s |
| IQ4_XS | 4.4 GB | ! Runs, barely | ~7.9 tokens/s |
| Q4_0 | 4.7 GB | ! Runs, barely | ~7.4 tokens/s |
| Q4_K_S | 4.7 GB | ! Runs, barely | ~7.4 tokens/s |
| Q4_K_M | 4.9 GB | ! Runs, barely | ~7.1 tokens/s |
| Q5_K_M | 5.7 GB | ! Runs, barely | ~6.1 tokens/s |
| Q6_K | 6.6 GB | ! Runs, barely | ~5.2 tokens/s |
| Q8_0 | 8.5 GB | ! Runs, barely | ~4.1 tokens/s |
Get it running in 3 steps
1
Install PocketPal from the Play Store — free, no account needed.
2
Download "Llama 3.1 8B Q3_K_M" in the in-app model hub — 4 GB, use Wi-Fi. Search Hugging Face for
bartowski/Meta-Llama-3.1-8B-Instruct-GGUF and pick the Q3_K_M file.3
Set context to 4K — expect ~8.6 tokens/s on this phone.
Related checks
MORE ON ROG PHONE 9 PRO
LLAMA 3.1 8B ON OTHER PHONES