qwen2.5:14b on A6000, A100, H100 — 2026-04-18
2026-04-18T02:31:20.000Z → 2026-04-18T02:32:59.025Z
On 2026-04-18, qwen2.5:14b ran across A6000, A100, H100. H100 finished first at 106.4 tok/s, 24× faster than A6000. The per-million-token cost: $6.50 (H100) vs $22.10 (A6000) — the surprise: the headline GPU H100 was also the cheapest per-million tokens, coming in 3.4× less than A6000.
Podium
H100
1st
thunder-h100
- peak tok/s
- 106.4
- avg tok/s
- 106.4
- $ / 1M tok
- $6.50
A100
2nd
thunder-a100
- peak tok/s
- 8.0
- avg tok/s
- 8.0
- $ / 1M tok
- $27.08
A6000
3rd
thunder-a6000
- peak tok/s
- 4.4
- avg tok/s
- 4.4
- $ / 1M tok
- $22.10