Explore benchmarks
Filter by model, hardware, workload and quantization. Every score links to the comparisons and experiments behind it.
Top Engines
| # | Engine | Speed score out of 100 |
|---|---|---|
| 1 | MLX LM†, an approximate comparison, see details | 97 /100 Score |
| ||
| 2 | Ollama†, an approximate comparison, see details | 89 /100 Score |
| ||
| 3 | llama.cpp†, an approximate comparison, see details | 86 /100 Score |
| ||
| Not ranked in this view | ||
| Not ranked | Magnitude | Too few results |
| Not ranked | Rapid-MLX | Too few results |
| Not ranked | LiteRT-LM | Too few results |
| Not ranked | uzu | Too few results |
| Not ranked | ExecuTorch | Too few results |
| Not ranked | ZML | Too few results |
details
Speed Decode speed: tokens generated per second.
Score: 100 is the best result in every comparison an engine entered; 80 is 80% of the best, on average. Only complete experiments with a passed output check count. How rankings work
The score combines counted comparisons. Select one comparison to see its measured values.
† MLX LM: approximate comparison. The experiments use different effective KV cache precision: litert-lm default, f16, magnitude default, model-dtype, uzu default. Its weights are 8-bit in another scheme: mlx-affine-8bit-g64, where most use Q8_0. The experiments use different effective KV cache precision: f16, magnitude default, model-dtype, zml default. The experiments use different effective KV cache precision: litert-lm default, f16, model-dtype. The experiments here use different 8-bit schemes: litert-int8, Q8_0, mlx-affine-8bit-g64. The experiments use different effective KV cache precision: bf16, fp32, litert-lm default, f16, magnitude default, model-dtype. Its weights are 4-bit in another scheme: mlx-affine-4bit-g64, where most use Q4_K_M. The experiments use different effective KV cache precision: f16, model-dtype. Its weights are 2-bit in another scheme: mlx-affine-2bit-g128, where most use Q2_0.
† Ollama: approximate comparison. The experiments use different effective KV cache precision: bf16, fp32, litert-lm default, f16, magnitude default, model-dtype, uzu default. The experiments use different effective KV cache precision: litert-lm default, f16, magnitude default, model-dtype, uzu default. The experiments use different effective KV cache precision: f16, magnitude default, model-dtype, zml default. The experiments use different effective KV cache precision: litert-lm default, f16, model-dtype. The experiments here use different 8-bit schemes: litert-int8, Q8_0, mlx-affine-8bit-g64. The experiments use different effective KV cache precision: bf16, fp32, litert-lm default, f16, magnitude default, model-dtype. The experiments use different effective KV cache precision: f16, model-dtype.
† llama.cpp: approximate comparison. The experiments use different effective KV cache precision: f16, uzu default. The experiments here use different 8-bit schemes: Q8_0, mirai-l. The experiments use different effective KV cache precision: bf16, fp32, litert-lm default, f16, magnitude default, model-dtype, uzu default. The experiments use different effective KV cache precision: litert-lm default, f16, magnitude default, model-dtype, uzu default. The experiments use different effective KV cache precision: f16, magnitude default, model-dtype, zml default. The experiments use different effective KV cache precision: litert-lm default, f16, model-dtype. The experiments here use different 8-bit schemes: litert-int8, Q8_0, mlx-affine-8bit-g64. The experiments use different effective KV cache precision: bf16, fp32, litert-lm default, f16, magnitude default, model-dtype. The experiments use different effective KV cache precision: f16, model-dtype. The experiments use different effective KV cache precision: fp32, litert-lm default, f16. The experiments here use different 4-bit schemes: 8da4w, litert-int4, Q4_K_M. The experiments use different effective KV cache precision: litert-lm default, f16. The experiments here use different 8-bit schemes: litert-int8, Q8_0.
Magnitude: Results in 6 of 16 comparisons; a rank needs 8.
Rapid-MLX: Results in 6 of 16 comparisons; a rank needs 8.
LiteRT-LM: Results in 5 of 16 comparisons; a rank needs 8.
uzu: Results in 4 of 16 comparisons; a rank needs 8.
ExecuTorch: Results in 2 of 16 comparisons; a rank needs 8.
ZML: Results in 2 of 16 comparisons; a rank needs 8.
Coverage
| Desktop | Mobile | |||
|---|---|---|---|---|
| M5 Max 64 GiB macOS 27 Apple M5 Max | OnePlus 13 16 GB Android 16 Qualcomm Snapdragon 8 Elite (SM8750) | Pixel 8 8 GB Android 16 Google Tensor G3 | iPhone 12 4 GB iOS 27 Apple A14 Bionic | |
| llama.cpp | 8511/11 out of 100, in 11 of 11 comparisons | 1003/3 out of 100, in 3 of 3 comparisons | 811/1 out of 100, in 1 of 1 comparisons | 691/1 out of 100, in 1 of 1 comparisons |
| LiteRT-LM | 2/11Not ranked. Results in 2 of 11 comparisons; a rank needs 6. | 812/3 out of 100, in 2 of 3 comparisons | 1001/1 out of 100, in 1 of 1 comparisons | Not measured on this device. |
| uzu | 3/11Not ranked. Results in 3 of 11 comparisons; a rank needs 6. | Not measured on this device. | Not measured on this device. | 1001/1 out of 100, in 1 of 1 comparisons |
| Magnitude | 936/11 out of 100, in 6 of 11 comparisons | Not measured on this device. | Not measured on this device. | Not measured on this device. |
| MLX LM | 979/11 out of 100, in 9 of 11 comparisons | Not measured on this device. | Not measured on this device. | Not measured on this device. |
| Ollama | 8911/11 out of 100, in 11 of 11 comparisons | Not measured on this device. | Not measured on this device. | Not measured on this device. |
| Rapid-MLX | 976/11 out of 100, in 6 of 11 comparisons | Not measured on this device. | Not measured on this device. | Not measured on this device. |
| ExecuTorch | 1/11Not ranked. Results in 1 of 11 comparisons; a rank needs 6. | 1/3Not ranked. Results in 1 of 3 comparisons; a rank needs 2. | Not counted Partial: 2 of 6 planned trials were valid. | Not counted Partial: 4 of 6 planned trials were valid. |
| ZML | 2/11Not ranked. Results in 2 of 11 comparisons; a rank needs 6. | Not measured on this device. | Not measured on this device. | Not measured on this device. |
Speed
| Rank | Engine | Decode and TTFT |
|---|---|---|
| 1 | uzu†, approximate comparison, see details | 70.2 tok/s 2,356 ms TTFT |
| 2 | llama.cpp†, approximate comparison, see details | 48.3 tok/s 2,210 ms TTFT |
details
LFM2.5 350M on iPhone 12 · 4 GB · 8-bit · Archive 2048 input / 128 output · iOS 27
Medians with the model warmed; model loading is excluded. Each table uses one matched comparison.
uzu: † Approximate comparison. The experiments use different effective KV cache precision: f16, uzu default. The experiments here use different 8-bit schemes: Q8_0, mirai-l.
llama.cpp: † Approximate comparison. The experiments use different effective KV cache precision: f16, uzu default. The experiments here use different 8-bit schemes: Q8_0, mirai-l.
Memory usage by engine
| Rank | Engine | Memory |
|---|---|---|
| Not ranked | llama.cpp | Not measured |
| Not ranked | uzu | Not measured |
details
LFM2.5 350M on iPhone 12 · 4 GB · 8-bit · Archive 2048 input / 128 output · iOS 27
Memory ranks require comparable, calibrated measurements. Unavailable measurements remain unranked.
llama.cpp: No peak memory measured. † Approximate comparison. The experiments use different effective KV cache precision: f16, uzu default. The experiments here use different 8-bit schemes: Q8_0, mirai-l.
uzu: No peak memory measured. † Approximate comparison. The experiments use different effective KV cache precision: f16, uzu default. The experiments here use different 8-bit schemes: Q8_0, mirai-l.
Every comparison
| Comparison | MLX LM | Ollama | llama.cpp | Magnitude | Rapid-MLX | LiteRT-LM | uzu | ExecuTorch | ZML |
|---|---|---|---|---|---|---|---|---|---|
| LFM2.5 350M on iPhone 12 · 4 GB 8-bit · Archive 2048 input / 128 output · iOS 27 | Not run | Not run | 69†, approximate | Not run | Not run | Not run | 100†, approximate | Not run | Not run |
| LFM2.5 1.2B Instruct on M5 Max · 64 GiB 4-bit · Archive 2048 input / 128 output · macOS 27 | Not counted | 84†, approximate | 83†, approximate | 85†, approximate | Not counted | Not counted | 100†, approximate | 86†, approximate | Not run |
| LFM2.5 1.2B Instruct on M5 Max · 64 GiB 8-bit · Archive 2048 input / 128 output · macOS 27 | 92†, approximate | 87†, approximate | 87†, approximate | 89†, approximate | 92†, approximate | 32†, approximate | 100†, approximate | Not run | Not run |
| LFM2.5 1.2B Instruct on M5 Max · 64 GiB bf16 · Archive 2048 input / 128 output · macOS 27 | 99†, approximate | 99†, approximate | 92†, approximate | 100†, approximate | 98†, approximate | Not run | Not run | Not run | 76†, approximate |
| LFM2.5 2.6B on M5 Max · 64 GiB 8-bit · Archive 2048 input / 128 output · macOS 27 | 100†, approximate | 96†, approximate | 89†, approximate | Unsupported | 100†, approximate | Not counted | Unsupported | Not run | Not run |
| LFM2.5 2.6B on M5 Max · 64 GiB bf16 · Archive 2048 input / 128 output · macOS 27 | Not run | 100 | 97 | Unsupported | Not run | Not run | Not run | Not run | Unsupported |
| LFM2.5 350M on M5 Max · 64 GiB 4-bit · Archive 2048 input / 128 output · macOS 27 | 95†, approximate | 74†, approximate | 69†, approximate | 93†, approximate | 100†, approximate | Not counted | Failed | Not counted | Not run |
| LFM2.5 350M on M5 Max · 64 GiB 8-bit · Archive 2048 input / 128 output · macOS 27 | 90†, approximate | 77†, approximate | 76†, approximate | 96†, approximate | 92†, approximate | 26†, approximate | 100†, approximate | Not run | Not run |
| LFM2.5 350M on M5 Max · 64 GiB bf16 · Archive 2048 input / 128 output · macOS 27 | 99†, approximate | 88†, approximate | 83†, approximate | 96†, approximate | 100†, approximate | Not run | Not run | Not run | 49†, approximate |
| Qwen3.8 27B on M5 Max · 64 GiB 4-bit · Archive 2048 input / 128 output · macOS 27 | 100†, approximate | 78†, approximate | 76†, approximate | Unsupported | Not run | Not run | Unsupported | Not run | Not run |
| Qwen3.8 27B on M5 Max · 64 GiB 8-bit · Archive 2048 input / 128 output · macOS 27 | 97†, approximate | 100†, approximate | 98†, approximate | Unsupported | Not run | Not run | Unsupported | Not run | Not run |
| Ternary Bonsai 2 27B on M5 Max · 64 GiB 2-bit · Archive 2048 input / 128 output · macOS 27 | 97†, approximate | 100†, approximate | 96†, approximate | Unsupported | Not run | Not run | Not run | Not run | Not run |
| LFM2.5 1.2B Instruct on OnePlus 13 · 16 GB 4-bit · Archive 2048 input / 128 output · Android 16 | Not run | Not run | 99†, approximate | Not run | Not run | Not counted | Not run | 100†, approximate | Not run |
| LFM2.5 1.2B Instruct on OnePlus 13 · 16 GB 8-bit · Archive 2048 input / 128 output · Android 16 | Not run | Not run | 100†, approximate | Not run | Not run | 87†, approximate | Not run | Not run | Not run |
| LFM2.5 350M on OnePlus 13 · 16 GB 8-bit · Archive 2048 input / 128 output · Android 16 | Not run | Not run | 100†, approximate | Not run | Not run | 76†, approximate | Not run | Not run | Not run |
| LFM2.5 350M on Pixel 8 · 8 GB 8-bit · Archive 2048 input / 128 output · Android 16 | Not run | Not run | 81†, approximate | Not run | Not run | 100†, approximate | Not run | Not run | Not run |
| LFM2.5 1.2B Instruct on iPhone 12 · 4 GB 4-bit · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Measured†, approximate | Not run | Not run | Not run | Not counted | Not counted | Not run |
| LFM2.5 1.2B Instruct on iPhone 12 · 4 GB 8-bit · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Measured†, approximate | Not run | Not run | Not run | Not counted | Not run | Not run |
| LFM2.5 1.2B Instruct on iPhone 12 · 4 GB bf16 · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Unsupported | Not run | Not run | Not run | Not run | Not run | Not run |
| LFM2.5 2.6B on iPhone 12 · 4 GB 4-bit · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Not counted | Not run | Not run | Not run | Unsupported | Not run | Not run |
| LFM2.5 2.6B on iPhone 12 · 4 GB 8-bit · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Unsupported | Not run | Not run | Not run | Unsupported | Not run | Not run |
| LFM2.5 350M on iPhone 12 · 4 GB 4-bit · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Measured†, approximate | Not run | Not run | Not run | Failed | Not counted | Not run |
| LFM2.5 350M on iPhone 12 · 4 GB bf16 · Archive 2048 input / 128 output · iOS 27 Not counted: fewer than two scored engines | Not run | Not run | Measured | Not run | Not run | Not run | Not run | Not run | Not run |
| LFM2.5 2.6B on M5 Max · 64 GiB 4-bit · Archive 2048 input / 128 output · macOS 27 Not counted: fewer than two scored engines | Not counted | Not counted | Not counted | Unsupported | Not counted | Measured†, approximate | Unsupported | Not run | Not run |
| LFM2.5 1.2B Instruct on OnePlus 13 · 16 GB bf16 · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Failed | Not run | Not run | Not run | Not run | Not run | Not run |
| LFM2.5 2.6B on OnePlus 13 · 16 GB 4-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Not counted | Not run | Not run | Measured†, approximate | Not run | Not run | Not run |
| LFM2.5 2.6B on OnePlus 13 · 16 GB 8-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Measured†, approximate | Not run | Not run | Not counted | Not run | Not run | Not run |
| LFM2.5 350M on OnePlus 13 · 16 GB 4-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Measured†, approximate | Not run | Not run | Not counted | Not run | Not counted | Not run |
| LFM2.5 350M on OnePlus 13 · 16 GB bf16 · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Measured | Not run | Not run | Not run | Not run | Not run | Not run |
| LFM2.5 1.2B Instruct on Pixel 8 · 8 GB 4-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Not counted | Not run | Not run | Not counted | Not run | Not counted | Not run |
| LFM2.5 1.2B Instruct on Pixel 8 · 8 GB 8-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Measured†, approximate | Not run | Not run | Not counted | Not run | Not run | Not run |
| LFM2.5 1.2B Instruct on Pixel 8 · 8 GB bf16 · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Failed | Not run | Not run | Not run | Not run | Not run | Not run |
| LFM2.5 2.6B on Pixel 8 · 8 GB 4-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Failed | Not run | Not run | Not counted | Not run | Not run | Not run |
| LFM2.5 2.6B on Pixel 8 · 8 GB 8-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Failed | Not run | Not run | Failed | Not run | Not run | Not run |
| LFM2.5 350M on Pixel 8 · 8 GB 4-bit · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Not counted | Not run | Not run | Not counted | Not run | Not counted | Not run |
| LFM2.5 350M on Pixel 8 · 8 GB bf16 · Archive 2048 input / 128 output · Android 16 Not counted: fewer than two scored engines | Not run | Not run | Failed | Not run | Not run | Not run | Not run | Not run | Not run |
Each engine’s Speed score in each matched comparison, out of 100, or why it has none. How rankings work