Qwen2.5-7B-Instruct-4bit
3 workload results across 2 hardware configurations.
Fastest local config
139.6 decode tok/s
on M3 Ultra (60-core GPU) + 96GB unified via mlx (4bit) — see full run
Local runs (3 runs)
Runs from contributors' own machines via MLX, llama.cpp, vLLM, exllamav2, or ollama. Signed on the submitter's hardware.
M3 Pro (18-core GPU) + 36GB unified
| Workload | Backend | Quant | decode tok/s | prefill tok/s | TTFT | Run |
|---|---|---|---|---|---|---|
| chat-short | mlx@0.31.3 | 4bit | 30.52tok/s | 161.9tok/s | 809ms | r_llzv_g-ymaf |
| chat-short | mlx@0.31.3 | 4bit | 15.67tok/s | 54.34tok/s | 2,411ms | r_e3t93rscswq |
M3 Ultra (60-core GPU) + 96GB unified
| Workload | Backend | Quant | decode tok/s | prefill tok/s | TTFT | Run |
|---|---|---|---|---|---|---|
| chat-short | mlx@0.31.3 | 4bit | 139.6tok/s | 190.0tok/s | 689ms | r_5r6rhiynenc |