Verified November 2026

Home › Compare › Mistral Large 4 vs Mistral Small 4

Mistral Large 4 vs Mistral Small 4

A straight cost comparison of Mistral Large 4 (Mistral AI) and Mistral Small 4 (Mistral AI), per 1,000,000 tokens, verified November 2026.

$0.70Mistral Large 4 input / 1M
$0.15Mistral Small 4 input / 1M
76%Blended price gap
Mistral Small 4Cheaper blended

Rate-by-rate comparison

USD per 1M tokens · November 2026
MetricMistral Large 4Mistral Small 4Winner
Input / 1M$0.70$0.15Mistral Small 4
Cached input / 1M$0.07$0.015Mistral Small 4
Cache write / 1M——Tie
Output / 1M$2.10$0.60Mistral Small 4
Context window1,000,000256,000Mistral Large 4
Batch APIYesYesTie
Long-context tierNoneNone—

For per-token metrics, the lower number wins. For context window and batch support, the higher or broader option wins. A dash means the provider does not publish that rate: a missing cache-write fee means writes are not billed per token, while a missing cached-input rate means cache hits bill at the full input rate.

Blended cost: 1M input + 250K output

ModelBlended costDifference
Mistral Large 4$1.23+$0.925
Mistral Small 4$0.30Cheaper

Ten workloads, two bills

The same two models, priced across ten workloads that differ only in their input:output ratio. Nothing else changes — no retries, no system prompt, no middleware.

WorkloadMistral Large 4Mistral Small 4Winner
Classification
100K reqs · 500 in / 10 out
$37.10$8.10Mistral Small 4
Data extraction
10K reqs · 2K in / 200 out
$18.20$4.20Mistral Small 4
Support chat
1K reqs · 4K in / 600 out
$4.06$0.96Mistral Small 4
Code review
2K reqs · 8K in / 1.5K out
$17.50$4.20Mistral Small 4
Summarisation
1K reqs · 20K in / 1K out
$16.10$3.60Mistral Small 4
Agent loop
5K reqs · 12K in / 2K out
$63$15Mistral Small 4
Translation
500 reqs · 15K in / 15K out
$21$5.63Mistral Small 4
Long-document QA
200 reqs · 120K in / 800 out
$17.14$3.70Mistral Small 4
Cached RAG
1K reqs · 50K in (45K cached) / 1K out
$8.78$2.03Mistral Small 4
Nightly batch
10K reqs · 10K in / 2K out
$56$13.50Mistral Small 4

Mistral Large 4 wins 0 of 10, Mistral Small 4 wins 10 of 10. The winner is set by the token mix, not by the headline rate.

Where the two bills cross over

Mistral Small 4 is cheaper on both sides of the mix: $0.15 vs $0.70 per million input tokens, and $0.60 vs $2.10 per million output. No input:output ratio reverses that, so on price alone the other model never becomes the better choice — it has to win on context, latency or output quality.

Cost at four scales

The same support-chat workload (4K input, 600 output per request) at four monthly volumes.

Requests / monthMistral Large 4Mistral Small 4Difference
100$0.406$0.096Mistral Small 4 cheaper by $0.31
1,000$4.06$0.96Mistral Small 4 cheaper by $3.10
10,000$40.60$9.60Mistral Small 4 cheaper by $31
100,000$406$96Mistral Small 4 cheaper by $310

What caching does to each bill

Caching only helps the part of the prompt that repeats. This is the support-chat workload (1,000 requests, 4,000 input tokens each) with an increasing share of that prompt served from a warm cache, written once on the first request.

Cached share of the promptMistral Large 4Mistral Small 4Winner
0%$4.06$0.96Mistral Small 4
50%$2.80$0.69Mistral Small 4
80%$2.05$0.528Mistral Small 4
95%$1.67$0.448Mistral Small 4

Both models publish a cached-read rate, so the discount is comparable: Mistral Large 4 reads cache at $0.07 per million, Mistral Small 4 at $0.015.

Which one should you pick?

Mistral Large 4 wins on none of the ten workloads above, while Mistral Small 4 wins on classification, data extraction, support chat, code review, summarisation, agent loop, translation, long-document qa, cached rag, nightly batch.

Beyond price, the two differ on input rate ($0.70 vs $0.15); output rate ($2.10 vs $0.60); cached-input rate ($0.07 vs $0.015); context window (1,000,000 vs 256,000).

Batch mode is available on Mistral Large 4 and Mistral Small 4, which halves the nightly job. Both publish cached-input discounts, so a stable system prompt is worth caching on either.

FAQ

Is Mistral Large 4 cheaper than Mistral Small 4?

On a blended 1M-input-plus-250K-output workload, Mistral Small 4 is cheaper by 76% ($0.30 vs $1.23). The gap changes for cache-heavy prompts and for batch jobs, as the workload table shows.

Which handles longer prompts, Mistral Large 4 or Mistral Small 4?

Mistral Small 4 has the larger context window (1,000,000 vs 256,000 tokens). Neither model publishes a long-context surcharge.

Do both models support prompt caching?

Mistral Large 4 bills cached reads at $0.07 per 1M; Mistral Small 4 at $0.015.

Related comparisons

Run your own token volumes