Ministral 14 B2512 vs Dolphin Mistral 24 B Venice Edition: cost comparison
Prices updated 2026-10-28 · computed from live provider data
Short answer: Ministral 14 B2512 is cheaper on a
3:1 input/output blend — $0.200 per million tokens versus
$0.375, a 47% difference.
Ministral 14 B2512 offers the larger context window at 262,144 tokens, against 128,000 for Dolphin Mistral 24 B Venice Edition.
Head to head
| Ministral 14 B2512 | Dolphin Mistral 24 B Venice Edition | |
|---|---|---|
| Input, per 1M tokens | $0.200 | $0.200 |
| Output, per 1M tokens | $0.200 | $0.900 |
| Blended 3:1, per 1M | $0.200 | $0.375 |
| Context window | 262,144 | 128,000 |
| Vendor | mistralai | cognitivecomputations |
Monthly cost at real volumes
| Workload (input / output) | Ministral 14 B2512 | Dolphin Mistral 24 B Venice Edition |
|---|---|---|
| 1M in / 250K out | $0.250 | $0.425 |
| 10M in / 2.5M out | $2.50 | $4.25 |
| 100M in / 25M out | $25.00 | $42.50 |
Which should you pick?
Pick Ministral 14 B2512 when you need the longer context window (262,144 vs 128,000 tokens). Pick Ministral 14 B2512 when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.