OverpayingForAIPricing desk

Lesson 2 of 8 · 10 min read · Beginner

Mistral models explained: Large, Medium, Small, Ministral and the specialists

A map of the Mistral model families, what each one is for, the context windows, and live prices so you can pick the cheapest model that will do the job.

In this lesson you will

  • Recognise the Large, Medium, Small, Ministral, Codestral, Devstral, Voxtral and Nemo families by name
  • Pick a default model for chat, coding, bulk processing and audio without overspending
  • Read a Mistral model id like mistral-small-2603 and know what the number means

Mistral names its models by size band, not by generation. Large is the flagship, Medium the workhorse, Small the value tier, and Ministral the edge-sized models built to run on laptops and phones. Around those sit specialists: Codestral for code completion, Devstral for agentic coding, Voxtral for speech, and Nemo, an older small model that is still one of the cheapest things you can rent.

Mistral AI: input and output price per 1M tokensLive from the OverpayingForAI catalogue · last verified 2026-09-04 · sorted by output priceInputOutputMistral Medium 3.5262K context$1.5$7.5Devstral 2 2512262K context$0.4$2Mistral Medium 3.1131K context$0.4$2Mistral Large 3 2512262K context$0.5$1.5Codestral 2508256K context$0.3$0.9Mistral Small 4262K context$0.15$0.6Mistral Small 3.2 24B131K context$0.075$0.2Ministral 3 8B 2512262K context$0.15$0.15Batch, fast-mode and free variants are excluded. Prices change; the figure re-draws from the catalogue on every build.
Figure 1.Live from our catalogue. Output tokens usually cost several times more than input tokens, so a chatty model costs more than a terse one at the same list price.

Reading the model id

An id like mistral-small-2603 means the Small family, snapshot from March 2026 (YYMM). ministral-8b-2512 is the 8-billion-parameter Ministral released December 2025. The -latest aliases (for example mistral-small-latest) always point at the newest snapshot, which is convenient for prototypes and dangerous for production, because the price and behaviour can change under you. Pin a dated id once something ships.

Mistral families at the time of writing (prices in $ per 1M tokens from our catalogue; check the live ladder above for changes)
FamilyCurrent modelInputOutputContextPick it for
LargeMistral Large 3 2512$0.50$1.50262KBest general quality Mistral sells; long documents, hard questions
MediumMistral Medium 3.5$1.50$7.50262KNewest mid-tier; check it earns its premium over Large 3 on your tasks
MediumMistral Medium 3.1$0.40$2.00131KPrevious mid-tier, still a strong default for chat and drafting
SmallMistral Small 4$0.15$0.60262KEveryday summarising, extraction, classification, open-weight
SmallMistral Small 3.2 24B$0.075$0.20131KSame job, older snapshot, even cheaper
MinistralMinistral 3 8B 2512$0.15$0.15262KBulk jobs where output volume is high; runs on a laptop
MinistralMinistral 3 3B 2512$0.10$0.10131KTagging, routing, on-device; cheapest new Ministral
CodestralCodestral 2508$0.30$0.90256KAutocomplete and fill-in-the-middle code generation
DevstralDevstral 2 2512$0.40$2.00262KAgentic coding: multi-file edits, tool use
VoxtralVoxtral Small 24B 2507$0.10$0.3032KSpeech understanding and transcription-style tasks
NemoMistral Nemo$0.019$0.03131KThe cheapest row in the catalogue; simple high-volume text

Which one should you actually pick?

  • Default for most work: Mistral Small 4. It has the full 262K context, it is open-weight, and it is a fraction of the Large price.
  • When Small gets it wrong: step up to Mistral Large 3 or Medium 3.1, then test Medium 3.5 only if the others fail and the task is worth five times the output price.
  • High-volume, short outputs: Ministral 3 8B or 3B. Symmetric pricing (input and output cost the same) makes them predictable for classification and routing.
  • Code in an editor: Codestral. Code agents that edit whole repos: Devstral 2.
  • Audio in, text out: Voxtral Small. Watch the 32K context.
  • If you are truly cost-bound: Mistral Nemo is one of the cheapest hosted models anywhere. Quality is dated; try it before you scoff.

Context windows

Most current Mistral models sit at 131K or 262K tokens of context, which is roughly a 200- to 400-page book. That is plenty for almost any single document. The catch is that context is not free: every token you put in the window is billed on every request. A 200K-token prompt on Large 3 costs about $0.10 per call before the answer. Pass what the model needs, not the whole archive.

Knowledge check

You need to classify 2 million short support tickets into 12 categories. Which Mistral model is the sensible first choice?

Lesson FAQ

What does the number in a Mistral model id mean?

It is a YYMM date stamp. mistral-small-2603 is the Small snapshot from March 2026; ministral-8b-2512 is from December 2025. The -latest aliases follow the newest snapshot.

Which Mistral models are open-weight?

Most of the Small, Ministral, Nemo, Codestral and Devstral lines are published under Apache 2.0. Check each model card on docs.mistral.ai, because licences differ by release.

Is Mistral Large the most expensive Mistral model?

Not in our catalogue at the time of writing. Mistral Medium 3.5 lists higher per token than Mistral Large 3. Always check the live price ladder rather than trusting the name.

Finished reading?

Mark it done to track your progress through the course.

Compare, calculate, decide — for Mistral (Le Chat)

If our calculators helped you cut down on hidden AI wallet leaks, consider buying us a coffee. A tiny fraction of your savings keeps our pricing indexes updated daily.

Not sure which AI is cheapest for your use case? Find out in 30 seconds — no signup required.

AI cost intelligence

Stop overpaying for AI tools

Join the OverpayingForAI list for pricing updates, cheaper alternatives, and practical buying guidance.

Now tracking 50+ AI tools, models, platforms, subscriptions, coding tools, and automation products.

We use your email only for OverpayingForAI updates. Unsubscribe anytime.