Explore this comparison’s ranking

Comparison Not counted

Which engine runs LFM2.5 1.2B Instruct best on iPhone 12 · 4 GB?

Device
iPhone 12 · 4 GB (iPhones)
OS
iOS 27
Model
LFM2.5 1.2B Instruct, 1.2B parameters
Weights
4-bit
Workload
Archive 256 input / 128 output: 256 tokens in, 128 out
Decoding
Greedy, batch 1, context 4,096
Power
Plugged in

Memory: Peak memory while generating: the less, the better.

Not counted: fewer than two scored engines. A comparison counts once two engines have scores for this objective. Its experiments are below, with their own numbers. How rankings work

Every experiment

Medians over valid trials, with the lowest and highest beneath. Eligible experiments are complete and passed their output check. The ranking above selects each engine's configuration for this objective when its required metrics are measured.

Every experiment in this comparison
ExperimentRanking eligibilityDecode tok/s, higher is betterFirst token ms, lower is betterMemory GiB, lower is betterPrefill tok/s, higher is betterValid trials
ExecuTorch 1.6.0 · CPU · 8da4w Eligible, approximately: The experiments use different effective KV cache precision: fp32, f16, uzu default. The experiments here use different 4-bit schemes: 8da4w, Q4_K_M, mirai-m.42.5 range40.4–42.82,033 range1,990–2,372Not measured125.9 range107.9–128.66of 6 planned
llama.cpp b12110 · Metal · Q4_K_M Eligible, approximately: The experiments use different effective KV cache precision: fp32, f16, uzu default. The experiments here use different 4-bit schemes: 8da4w, Q4_K_M, mirai-m.37.6 range37.5–37.7987 range984–993Not measured259.7 range257.9–260.46of 6 planned
uzu 0.6.17 · Metal · mirai-m Eligible, approximately: The experiments use different effective KV cache precision: fp32, f16, uzu default. The experiments here use different 4-bit schemes: 8da4w, Q4_K_M, mirai-m.45.4 range44.4–46.01,073 range1,055–1,117Not measured257.7 range257.1–259.16of 6 planned

Every trial, metric by metric

The line runs from an experiment's lowest to its highest valid trial, the dot is the median and each tick is one trial. Every scale starts at zero.

Caveats

  • Publisher cards identify the base model but do not always pin the exact export source checkpoint revision.
  • Engines may differ in quantization scheme, effective KV cache precision, kernels, and activations.

About this comparison

Worked out from its experiments: every experiment here shares the device, the OS major version, the power state, the base model, the quantization class and the workload. Measured under methodology v0.6. Comparison id glenn-iphone-12--ios-27--lfm2.5-1.2b-instruct--4-bit--archive-256-128.