Gemma 4 26 B vs GLM 5.2: cost comparison
Prices updated 2026-10-28 · computed from live provider data
Short answer: Gemma 4 26 B is cheaper on a
3:1 input/output blend — $0.064 per million tokens versus
$1.78, a 96% difference.
GLM 5.2 offers the larger context window at 1,048,576 tokens, against 131,072 for Gemma 4 26 B.
Head to head
| Gemma 4 26 B | GLM 5.2 | |
|---|---|---|
| Input, per 1M tokens | $0.030 | $0.034 |
| Output, per 1M tokens | $0.165 | $7.00 |
| Blended 3:1, per 1M | $0.064 | $1.78 |
| Context window | 131,072 | 1,048,576 |
| Vendor | darkbloom | z-ai |
Monthly cost at real volumes
| Workload (input / output) | Gemma 4 26 B | GLM 5.2 |
|---|---|---|
| 1M in / 250K out | $0.071 | $1.78 |
| 10M in / 2.5M out | $0.713 | $17.84 |
| 100M in / 25M out | $7.12 | $178 |
Which should you pick?
Pick GLM 5.2 when you need the longer context window (1,048,576 vs 131,072 tokens). Pick Gemma 4 26 B when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.