Gemma 4 26 B A4 B vs Llama 3.2.3 B Instruct: cost comparison
Prices updated 2026-11-05 · computed from live provider data
Short answer: Gemma 4 26 B A4 B is cheaper on a
3:1 input/output blend — $0.110 per million tokens versus
$0.122, a 10% difference.
Gemma 4 26 B A4 B offers the larger context window at 262,144 tokens, against 131,072 for Llama 3.2.3 B Instruct.
Head to head
| Gemma 4 26 B A4 B | Llama 3.2.3 B Instruct | |
|---|---|---|
| Input, per 1M tokens | $0.050 | $0.051 |
| Output, per 1M tokens | $0.290 | $0.335 |
| Blended 3:1, per 1M | $0.110 | $0.122 |
| Context window | 262,144 | 131,072 |
| Vendor | empiriolabs | meta-llama |
Monthly cost at real volumes
| Workload (input / output) | Gemma 4 26 B A4 B | Llama 3.2.3 B Instruct |
|---|---|---|
| 1M in / 250K out | $0.122 | $0.135 |
| 10M in / 2.5M out | $1.23 | $1.35 |
| 100M in / 25M out | $12.25 | $13.46 |
Which should you pick?
Pick Gemma 4 26 B A4 B when you need the longer context window (262,144 vs 131,072 tokens). Pick Gemma 4 26 B A4 B when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.