Llama 4 Maverick 17 B128 E Instruct Fp 8 vs Llama 4 Scout 17 B16 E Instruct: cost comparison
Prices updated 2026-10-27 · computed from live provider data
Short answer: Llama 4 Scout 17 B16 E Instruct is cheaper on a
3:1 input/output blend — $0.062 per million tokens versus
$0.062, a 0% difference.
Llama 4 Maverick 17 B128 E Instruct Fp 8 offers the larger context window at 131,072 tokens, against 16,384 for Llama 4 Scout 17 B16 E Instruct.
Head to head
| Llama 4 Maverick 17 B128 E Instruct Fp 8 | Llama 4 Scout 17 B16 E Instruct | |
|---|---|---|
| Input, per 1M tokens | $0.050 | $0.050 |
| Output, per 1M tokens | $0.100 | $0.100 |
| Blended 3:1, per 1M | $0.062 | $0.062 |
| Context window | 131,072 | 16,384 |
| Vendor | lambda_ai | lambda_ai |
Monthly cost at real volumes
| Workload (input / output) | Llama 4 Maverick 17 B128 E Instruct Fp 8 | Llama 4 Scout 17 B16 E Instruct |
|---|---|---|
| 1M in / 250K out | $0.075 | $0.075 |
| 10M in / 2.5M out | $0.750 | $0.750 |
| 100M in / 25M out | $7.50 | $7.50 |
Which should you pick?
Pick Llama 4 Maverick 17 B128 E Instruct Fp 8 when you need the longer context window (131,072 vs 16,384 tokens). Pick Llama 4 Scout 17 B16 E Instruct when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.