Run Llama 3.2 1B locally on any laptop
Llama 3.2 1B Instruct is the smallest model in the catalog: a 0.8 GiB file that runs on any laptop, with or without a GPU. It is the model of the first-run recording on the main page: 68 seconds from an empty home folder to the first answer.
flelink run llama-3.2-1bWhat you need
Any machine: macOS, Linux or Windows, no GPU required.
Measured speed
| Recipe | Decode | Measured on |
|---|---|---|
| Llama 3.2 1B1B, a first try on any laptop | 91 tok/s | MacBook Air M2, 8 GB: the recording at the top of this page (3 Nov 2026) |
Does Llama 3.2 1B fit my card?
| Graphics | 32 GB RAM | 64 GB RAM | 128 GB RAM |
|---|---|---|---|
| 12 GB NVIDIA card | fits | fits | fits |
| 16 GB NVIDIA card | fits | fits | fits |
| 24 GB NVIDIA card | fits | fits | fits |
| 32 GB NVIDIA card | fits | fits | fits |
| Apple Silicon Mac | fits | fits | fits |
| No GPU | fits | fits | fits |
The verdicts are the rule FLE Link applies before it downloads anything (the rules in full). A speed is shown only for the hardware it was measured on. A size nobody measured yet? flelink bench --task pagoda --share posts yours to the benchmarks board.
Good to know
- Good for trying FLE Link out before a download of tens of gigabytes.
Sources
Install FLE Link
curl -fsSL https://flelink.run/install.sh | shWindows: irm https://flelink.run/install.ps1 | iex. Open source, MIT. No account, no telemetry. What FLE Link does →