Seed 2.0 Lite vs Qwen 3.6 27 B Fp 8: cost comparison
Prices updated 2026-11-05 · computed from live provider data
Short answer: Seed 2.0 Lite is cheaper on a
3:1 input/output blend — $0.858 per million tokens versus
$1.04, a 18% difference.
Qwen 3.6 27 B Fp 8 offers the larger context window at 262,144 tokens, against 229,096 for Seed 2.0 Lite.
Head to head
| Seed 2.0 Lite | Qwen 3.6 27 B Fp 8 | |
|---|---|---|
| Input, per 1M tokens | $0.310 | $0.320 |
| Output, per 1M tokens | $2.50 | $3.20 |
| Blended 3:1, per 1M | $0.858 | $1.04 |
| Context window | 229,096 | 262,144 |
| Vendor | empiriolabs | tensormesh |
Monthly cost at real volumes
| Workload (input / output) | Seed 2.0 Lite | Qwen 3.6 27 B Fp 8 |
|---|---|---|
| 1M in / 250K out | $0.935 | $1.12 |
| 10M in / 2.5M out | $9.35 | $11.20 |
| 100M in / 25M out | $93.50 | $112 |
Which should you pick?
Pick Qwen 3.6 27 B Fp 8 when you need the longer context window (262,144 vs 229,096 tokens). Pick Seed 2.0 Lite when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.