M5Stack AI Pyramid / Qwen2.5 0.5B
Can a M5Stack AI Pyramid run Qwen2.5 0.5B?
Yes, at a comfortable chat speed. Measured generation speed: 25.5–37.11 tok/s across 2 public runs.
Rule of thumb: 10+ tok/s feels like a chat; 4–10 is usable for assistants and automations; under 4 is batch-job territory. Speeds depend on quantization, runtime, context length and cooling. Check each source.
| Board | Quant | Runtime | Gen tok/s | Prompt tok/s | TTFT ms | Source |
|---|---|---|---|---|---|---|
| M5Stack AI Pyramid M5Stack AI Pyramid (AX8850) | W4A16 | StackFlow | 37.11 | – | 140.17 | source ↗ accessed 2026-10-17 |
| M5Stack AI Pyramid M5Stack AI Pyramid (AX8850) | W8A16 | StackFlow | 25.5 | – | 143.02 | source ↗ accessed 2026-10-17 |
Qwen2.5 0.5B on other boards
- M5Stack LLM8850 27.05 tok/s best
- Jetson Orin Nano Super 13.31 tok/s best
- M5Stack Module LLM 12.52 tok/s best
- Raspberry Pi 5 1.54 tok/s best
Buy a M5Stack AI Pyramid
- M5Stack AI Pyramid (AX8850): price not verified yet Check price on Amazon
Ran Qwen2.5 0.5B on a M5Stack AI Pyramid? Submit your numbers →