Mistral La PlateformeOpen-WeightVerified 2026-11-12
Mistral Large 4 API Pricing & Cost Rules
API Model ID: mistral-large-4 · Context Window: 1,000,000 tokens · Max Output: 262,144 tokens
Context Tiers & Token Rate Card (USD per 1M Tokens)
| Tier | Std Input | Std Output | Cache Read | Write (Short) | Write (Long) |
|---|---|---|---|---|---|
| Standard (<= 1M) | $0.7 | $2.1 | $0.07 | $0.7 | N/A |
Prompt Caching Rules
Supported: Yes (explicit)
Minimum Cacheable Prefix: 64 tokens
Hourly Cache Storage: $0
Uses prompt_cache_key; 64-token cache blocks (min 64 tokens; cached_tokens billed in multiples of 64). Read = 10% ($0.07/1M); zero write surcharge.
Batch API & Hosting Notes
Batch Supported: Yes (50% discount · 24h default (<7d max))
Caching in Batch: Disabled in Batch mode
50% discount available on all Mistral models ($0.35/$1.05). Docs do not specify batch+cache discount stacking.
1.05T total / 52B active MoE open-weight model (Mistral Ecosystem Open Weight License).