How fast does Gemma 3 12B (Vision) run on iPhone?
Gemma 3 12B (Vision) hasn't been measured on the reference iPhone yet, so there's no speed figure for it here.
You can measure it on your own phone with Model Benchmark in the app. It runs the same protocol as the numbers on the benchmarks page.
How much memory and storage does Gemma 3 12B (Vision) need?
The catalog budgets 8.4 GB of RAM and 7.5 GB of storage for the 4bit build of Gemma 3 12B (Vision).
Its backend copies weights into the app's memory, so the app checks it against the per-app limit iOS enforces. On a 6 GB iPhone that limit is much lower than 6 GB.
Which iPhones can run Gemma 3 12B (Vision)?
Gemma 3 12B (Vision) needs more memory than iPhones give a single app today, so it's meant for iPads with more RAM.
| RAM | iPhones | Pocket AI Lab |
|---|---|---|
| 3–4 GB | iPhone 11, 11 Pro, 12, 12 mini, 13, 13 mini, SE (2nd and 3rd gen) | Not measured yet |
| 6 GB | iPhone 12 Pro, 13 Pro, 14, 14 Plus, 14 Pro, 15, 15 Plus | ✗ Too big |
| 8 GB | iPhone 15 Pro, 16, 16 Plus, 16 Pro, 16e, 17, 17e | ✗ Too big |
| 12 GB | iPhone 17 Pro, 17 Pro Max, iPhone Air | ✗ Too big |
Based on the memory budgets the app measures on each class of iPhone (October 2026). Before downloading, the app checks your exact phone and only offers what fits.
What can Gemma 3 12B (Vision) do in Pocket AI Lab?
It's a vision-language model, so besides text chat it can describe photos and frames sampled from a video you attach to a message.
Other Gemma models in the catalog
- Gemma 2 2B, MLX, 4bit, 30.3 tok/s
- Gemma 3 4B (Vision), MLX, 4bit, 21.8 tok/s (throttled)
- Gemma 3 4B (GGUF, Vision), llama.cpp, Q4_K_M, 20.7 tok/s (throttled)
- Gemma 4 E2B (GGUF, Vision), llama.cpp, Q4_K_M, 35.0 tok/s
Gemma 4 on iPhone compares the newer generation with this one and shows which Gemma 4 sizes fit each iPhone.
How do I install Gemma 3 12B (Vision)?
Open the Models tab in Pocket AI Lab and tap Gemma 3 12B (Vision). The app checks it against your phone's memory, downloads 7.5 GB from Hugging Face and it's ready for offline chat.