Compare
Loading community benchmarks…
Loading community benchmarks…
Put two chips, runtimes, or runtime versions side by side. Everything else is held fixed or called out, so a ratio is worth exactly what the matched runs say.
Ratios read side A relative to side B. Geometric mean over 2 like-for-like pairs.
| Facet | Apple M5 Pro | Apple M1 Max |
|---|---|---|
| Model | Qwen3-4B-Instruct-2507 | Qwen3-4B-Instruct-2507-Q4, Qwen3-4B-Instruct-2507-Q8 |
| Runtime | BaseRT, llama.cpp | BaseRT |
| Runtime version | BaseRT 0.2.4, llama.cpp b10809 (5266f24da) | BaseRT 0.2.6, BaseRT 0.3.0 |
| Quantisation | Q4, Q4_K_M, Q8, Q8_0 | Q4, Q8 |
| Backend | BLAS + Metal, Metal | Metal |
| Conditioning | runtime_native_warmup, warmup_only | warmup_only |
| Decode workload | TG128 | TG128 @ 1 ctx |
| Harness schema | basert-benchmark-harness/1, computearena-measurements/1 | basert-benchmark-harness/1 |
Each row is a configuration present on both sides. Values are per-cell medians; ratios read Apple M5 Pro relative to Apple M1 Max.
| Configuration | Decode A | Decode B | Ratio | Prefill A | Prefill B | Ratio | Runs A / B |
|---|---|---|---|---|---|---|---|
Qwen3-4B-Instruct-2507 BaseRTQ4 | 89.9 | 90.8 | 0.99× | 3,345 | 922 | 3.63× | 2 / 2 |
Qwen3-4B-Instruct-2507 BaseRTQ8 | 63.5 | 71.8 | 0.89× | 3,402 | 924 | 3.68× | 2 / 2 |
Top 7 of 7 runs by decode throughput.
| Model / format | Chip / backend | Decode | Prefill | Contributor | Date | Report |
|---|---|---|---|---|---|---|
Qwen3-4B-Instruct-2507 llama.cpp b10809 (5266f24da)Q4_K_M | Apple M5 Pro BLAS + Metal | 94.4 TG128 | 2,568 PP512 | arki05 | View Benchmark | |
Qwen3-4B-Instruct-2507 llama.cpp b10809 (5266f24da)Q4_K_M | Apple M5 Pro BLAS + Metal | 92.6 TG128 | 2,566 PP512 | arki05 | View Benchmark | |
basecompute/Qwen3-4B-Instruct-2507 BaseRT 0.2.4Q4 | Apple M5 Pro Metal | 90.4 TG128 | 3,331 PP512 | arki05 | View Benchmark | |
basecompute/Qwen3-4B-Instruct-2507 BaseRT 0.2.4Q4 | Apple M5 Pro Metal | 89.4 TG128 | 3,359 PP512 | arki05 | View Benchmark | |
basecompute/Qwen3-4B-Instruct-2507 BaseRT 0.2.4Q8 | Apple M5 Pro Metal | 64.3 TG128 | 3,398 PP512 | arki05 | View Benchmark | |
basecompute/Qwen3-4B-Instruct-2507 BaseRT 0.2.4Q8 | Apple M5 Pro Metal | 62.8 TG128 | 3,405 PP512 | arki05 | View Benchmark | |
Qwen3-4B-Instruct-2507 llama.cpp b10809 (5266f24da)Q8_0 | Apple M5 Pro BLAS + Metal | 60.8 TG128 | 2,686 PP512 | arki05 | View Benchmark |
Top 4 of 4 runs by decode throughput.
| Model / format | Chip / backend | Decode | Prefill | Contributor | Date | Report |
|---|---|---|---|---|---|---|
Qwen3-4B-Instruct-2507-Q4 BaseRT 0.2.6Q4 | Apple M1 Max Metal | 91.3 TG128 @ 1 ctx | 922 PP512 | basecompute | View Benchmark | |
Qwen3-4B-Instruct-2507-Q4 BaseRT 0.3.0Q4 | Apple M1 Max Metal | 90.4 TG128 @ 1 ctx | 922 PP512 | basecompute | View Benchmark | |
Qwen3-4B-Instruct-2507-Q8 BaseRT 0.2.6Q8 | Apple M1 Max Metal | 72.1 TG128 @ 1 ctx | 925 PP512 | basecompute | View Benchmark | |
Qwen3-4B-Instruct-2507-Q8 BaseRT 0.3.0Q8 | Apple M1 Max Metal | 71.4 TG128 @ 1 ctx | 923 PP512 | basecompute | View Benchmark |