Boards / M5Stack AI Pyramid

M5Stack AI Pyramid (AX8850): LLM benchmarks

M5Stack's AX8850-based AI box.

At a glance

  • Measured runs6
  • Models tested5
  • Fastest run37.11 tok/s
  • Qwen2.5 0.5B Instruct, W4A16, StackFlow · source ↗
  • Sources1

Where to buy

Every run on this board

ModelQuantRuntimeGen tok/sPrompt tok/sTTFT msSource
Qwen2.5 0.5B InstructW4A16StackFlow37.11–140.17source ↗
accessed 2026-10-17
InternVL3 1BW8A16StackFlow26.67–142.32source ↗
accessed 2026-10-17
Qwen2.5 0.5B InstructW8A16StackFlow25.5–143.02source ↗
accessed 2026-10-17
Qwen2.5 1.5B InstructW4A16StackFlow16.77–289.06source ↗
accessed 2026-10-17
Qwen3-VL 2B InstructW8A16StackFlow11.93–159.79source ↗
accessed 2026-10-17
Qwen2.5 3B InstructW4A16StackFlow9.46–550.3source ↗
accessed 2026-10-17

Can the M5Stack AI Pyramid run…

Can it run InternVL3 1B?Can it run Qwen2.5 0.5B?Can it run Qwen2.5 1.5B?Can it run Qwen2.5 3B?Can it run Qwen3-VL 2B?

Have this board? Submit your own run →

Get the monthly board report

New boards, new runtimes, new numbers. One email a month, no hype.