Mistral Medium 3
mistralai/mistral-medium-2505
spread (output, p90 ÷ p10)
— too few providers
$2.00 – $2.00 · 3 providers · max÷min 1.0×
Cheapest / median / spread use providers that serve the model on their own capacity: first-party lab, hyperscaler, inference host.
blended = 90.9% input × [(1 − 70% cache) × input + 70% cache × cache-read] + 9.1% × output + cache writes (30% of uncached input) where a write price is published
A provider that publishes no cache price is blended with cache share 0 (all input at the input price) and labelled “no cache price published” — never penalised against identical in/out prices.
| Provider | Input | Output | Cache rd | Blended |
|---|---|---|---|---|
Azure Cognitive Services hyperscaler | $0.40 | $2.00 | — | $0.55no cache price published |
Eden AI router / aggregator | $0.40 | $2.00 | — | $0.55no cache price published |
Merge Gateway router / aggregator | $0.40 | $2.00 | $0.04 | $0.32 |
Mistralfirst-party first-party lab | $0.40 | $2.00 | — | $0.55no cache price published |
NanoGPT router / aggregator | $0.40 | $2.00 | $0.20 | $0.42 |
Azure hyperscaler | $0.40 | $2.00 | — | $0.55no cache price published |
OpenRouter router / aggregator | $0.40 | $2.00 | $0.04 | $0.32 |
Free tiers (rate-limited) · 1
No payment required; rate limits and quotas apply (not tracked yet). Excluded from cheapest, median and spread.
- Nvidia mistralai/mistral-medium-3-instructNVIDIA developer account + API keyfree tier
Data: models.dev (MIT), snapshot 2026-09-20. Prices are vendor-published; verify before purchase. Methodology.