How much does the Mistral Small API cost?

Mistral Small lists at $0.15 per million input tokens and $0.6 per million output tokens as of August 2026. On a realistic task of 1,500 input and 500 output tokens across 3 calls, that is $0.00158 per task, or $1,575 per million tasks.

Is Mistral Small expensive compared to other models?

Mistral Small is the cheapest of the 10 models we track on cost per task. The most expensive tracked model, GPT-5.5, costs 42.9x more for the same work.

When does self-hosting beat the Mistral Small API?

Effectively never on commodity hardware: against a $2/hour GPU the break-even is 926,984 tasks per month, above the 788,400 tasks that box can physically serve, so the crossover is unreachable.

Model pricing · Mistral · value tier · August 2026

Mistral Small API pricing, in numbers that matter

List price, real cost per task, rank against the field, and the self-hosting break-even, all computed from our open dataset.

Input, $/1M tokens

$0.15

Output, $/1M tokens

$0.6

Cost per task

$0.00158

Cost rank

#1 of 10

The task profile behind every figure: 1,500 input and 500 output tokens per call, 3 calls per task. On that profile Mistral Small costs $0.00158 per task and $1,575 per million tasks. Output is billed at 4x the input rate, so despite the profile sending three times more input than output, output makes up 57% of the bill. Capping response length is the highest-return prompt-level optimisation on this model.

Mistral Small against the field

#	Model	$/task	vs Mistral Small	Head-to-head
1	Mistral Small	$0.00158	-	-
2	DeepSeek-V4-Flash	$0.00396	2.51x	compare
3	Gemini 2.5 Flash	$0.00510	3.24x	compare
4	Grok 4.3	$0.00938	5.95x	compare
5	GPT-5.4 mini	$0.01013	6.43x	compare
6	Claude Haiku 4.5	$0.01200	7.62x	compare
7	Claude Sonnet 5	$0.02400	15.2x	compare
8	Gemini 3.1 Pro	$0.02700	17.1x	compare
9	Claude Opus 4.8	$0.06000	38.1x	compare
10	GPT-5.5	$0.06750	42.9x	compare

Self-hosting break-even

Against a reference $2/hour dedicated instance ($1,460/month fixed, 788,400 tasks/month capacity at full utilisation), the break-even against Mistral Small is 926,984 tasks per month. That exceeds what the box can physically serve, so against this model that GPU never pays off at any volume. Work through your own numbers in the $/task calculator or the break-even guide.

Frequently asked

How much does the Mistral Small API cost?: Mistral Small lists at $0.15 per million input tokens and $0.6 per million output tokens as of August 2026. On a realistic task of 1,500 input and 500 output tokens across 3 calls, that is $0.00158 per task, or $1,575 per million tasks.
Is Mistral Small expensive compared to other models?: Mistral Small is the cheapest of the 10 models we track on cost per task. The most expensive tracked model, GPT-5.5, costs 42.9x more for the same work.
When does self-hosting beat the Mistral Small API?: Effectively never on commodity hardware: against a $2/hour GPU the break-even is 926,984 tasks per month, above the 788,400 tasks that box can physically serve, so the crossover is unreachable.