Qwen 3.5 27 B vs Qwen 3 235 B A22 B Instruct 2507 Fp 8: cost comparison
Prices updated 2026-11-05 · computed from live provider data
Short answer: Qwen 3 235 B A22 B Instruct 2507 Fp 8 is cheaper on a
3:1 input/output blend — $0.153 per million tokens versus
$0.236, a 35% difference.
Qwen 3 235 B A22 B Instruct 2507 Fp 8 offers the larger context window at 262,144 tokens, against 258,048 for Qwen 3.5 27 B.
Head to head
| Qwen 3.5 27 B | Qwen 3 235 B A22 B Instruct 2507 Fp 8 | |
|---|---|---|
| Input, per 1M tokens | $0.086 | $0.087 |
| Output, per 1M tokens | $0.688 | $0.350 |
| Blended 3:1, per 1M | $0.236 | $0.153 |
| Context window | 258,048 | 262,144 |
| Vendor | empiriolabs | gmi |
Monthly cost at real volumes
| Workload (input / output) | Qwen 3.5 27 B | Qwen 3 235 B A22 B Instruct 2507 Fp 8 |
|---|---|---|
| 1M in / 250K out | $0.258 | $0.175 |
| 10M in / 2.5M out | $2.58 | $1.75 |
| 100M in / 25M out | $25.80 | $17.50 |
Which should you pick?
Pick Qwen 3 235 B A22 B Instruct 2507 Fp 8 when you need the longer context window (262,144 vs 258,048 tokens). Pick Qwen 3 235 B A22 B Instruct 2507 Fp 8 when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.