Gemma 4 31 B vs Qwen 3.6 Flash Variant 1: cost comparison
Prices updated 2026-11-06 · computed from live provider data
Short answer: Gemma 4 31 B is cheaper on a
3:1 input/output blend — $0.240 per million tokens versus
$0.371, a 35% difference.
Qwen 3.6 Flash Variant 1 offers the larger context window at 983,616 tokens, against 128,000 for Gemma 4 31 B.
Head to head
| Gemma 4 31 B | Qwen 3.6 Flash Variant 1 | |
|---|---|---|
| Input, per 1M tokens | $0.164 | $0.165 |
| Output, per 1M tokens | $0.469 | $0.990 |
| Blended 3:1, per 1M | $0.240 | $0.371 |
| Context window | 128,000 | 983,616 |
| Vendor | kyma | empiriolabs |
Monthly cost at real volumes
| Workload (input / output) | Gemma 4 31 B | Qwen 3.6 Flash Variant 1 |
|---|---|---|
| 1M in / 250K out | $0.281 | $0.412 |
| 10M in / 2.5M out | $2.81 | $4.12 |
| 100M in / 25M out | $28.14 | $41.25 |
Which should you pick?
Pick Qwen 3.6 Flash Variant 1 when you need the longer context window (983,616 vs 128,000 tokens). Pick Gemma 4 31 B when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.