How fast does Mistral 7B v0.3 run on iPhone?
On an iPhone Air (A19, iOS 26.6.1), Mistral 7B v0.3 generated 9.1 tokens per second in Pocket AI Lab 1.1.0, averaged over 3 runs of 200 tokens (± 4.2).
On the standard llama-bench style probe it read a 512-token prompt at about 44 tok/s and generated 128 tokens at 4.4 tok/s. Cold load took 1.1 s and the process peaked at 4,671 MB. The phone heated up during this run and decode speed dropped 65% between repetitions, so on a cool phone it can be faster.
How much memory and storage does Mistral 7B v0.3 need?
The catalog budgets 4.5 GB of RAM and 4.4 GB of storage for the 4bit build of Mistral 7B v0.3. The measured peak footprint on iPhone Air was 4,671 MB.
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 Mistral 7B v0.3?
Mistral 7B v0.3 runs on iPhones with 8 GB of RAM or more.
| 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 | ✓ Runs |
| 12 GB | iPhone 17 Pro, 17 Pro Max, iPhone Air | ✓ Runs |
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 Mistral 7B v0.3 do in Pocket AI Lab?
It's a text model for chat, drafting, summaries and quick questions, fully offline once downloaded.
How did Mistral 7B v0.3 score on the quality suite?
It passed 61 of 118 tasks, a score of 51.7 out of 100 on suite v3.
Tasks are graded by deterministic checks (exact numbers, regex, word counts, valid JSON), with greedy decoding. See the method.
All runs of Mistral 7B v0.3
| Model | Gen tok/s | pp512 / tg128 | Load s | Peak MB | Score | Date |
|---|---|---|---|---|---|---|
| Mistral 7B v0.3 | 8.6 ± 3.6 | ~45 / 5.1 | 1.9 | 4,641 | 65.2 | 2026-09-03 |
| Mistral 7B v0.3 | 9.1 ± 4.2 | ~44 / 4.4 | 1.1 | 4,671 | 51.7 | 2026-10-22 |
How do I install Mistral 7B v0.3?
Open the Models tab in Pocket AI Lab and tap Mistral 7B v0.3. The app checks it against your phone's memory, downloads 4.4 GB from Hugging Face and it's ready for offline chat.