Mistral AI

• Mistral AI, European lab with an open-weight lineage and EU-resident hosting.
• - The credible European option: a Paris-based lab with EU data residency and on-premise licensing
• Releases capable models under Apache 2.0 alongside a commercial tier
• Rarely the capability leader
• Often the only shortlist entry that clears legal when jurisdiction, not benchmarks, is the blocker.
• Founded 2023.
• Late-stage private.
• Open source under Apache-2.0 (selected models).

• We track it in 1 comparison, LLM APIs.
• Every claim below is a cell in a table you can open and check rather than an impression.
• Across those rosters it sits against 17 other tools, and what follows is where it visibly separates from them.

Where it wins.

• $/M output, $1.5 /M tok, 2nd of 9.
• $/M input, $0.5 /M tok, 2nd of 9.
• $/M cache read, $0.05 /M tok, 2nd of 8.

• Each of those is ranked against the whole roster on its category page, not against a hand-picked subset.
• A first place here means first of everything we list.

Where it gives ground.

• Context window, 256,000 tokens, 10th of 12. Anthropic records 1,000,000 tokens.

• None of these disqualify it on their own.
• They are the fields to check against your own requirements before you commit, because they are the ones where a competitor genuinely does better.

• Provenance. 22 of 27 tracked fields carry a value for Mistral AI, and 12 of those cite a document you can open.
• Last verified 2026-07-30.
• Every figure keeps its own provenance, measured by us, claimed by the vendor, inferred, or community-reported, and we would rather print a dash than a guess.

• Its nearest neighbour in our data is Qwen.
• Mistral AI is ahead on Schema output (Yes against Partial) and Image input (Yes against Partial).
• Qwen takes Context window (1,000,000 tokens against 256,000 tokens) and Trains on your data (Partial against No).
• That pattern repeats across the rest of the roster, see Mistral AI alternatives for the other rivals, each compared the same way.

Founded

2023

Funding

Late-stage private

Licence

Apache-2.0 (selected models) (open source)

Fields we track

22 of 27

Last verified

2026-07-30

Rarely the capability leader, consistently the answer when jurisdiction is the blocker; check licences per model.

$/M output

$1.5/M tok2nd of 9

Vendor-claimedverified 2026-07-30source

$/M input

$0.5/M tok2nd of 9

Vendor-claimedverified 2026-07-30source

$/M cache read

$0.05/M tok2nd of 8

Vendor-claimedverified 2026-07-30source

Context window

256,000tokens10th of 12

Vendor-claimedverified 2026-07-30source

Tool use

●Yes1st of 18

Vendor-claimedverified 2026-07-30source

Schema output

●Yes1st of 18

Inferredverified 2026-01-15

Pinnable versions

●Yes1st of 18

Vendor-claimedverified 2026-01-15source

Trains on your data

○No7th of 18

Inferredverified 2026-01-15

Mistral AI is the right call in these situations, each one drawn from a field we actually record.

• EU data-residency requirements that rule out US-hosted inference.
• On-premise deployment with a commercial support contract.
• Teams that want an open-weight escape hatch under the hosted API.
• $/M output is your binding constraint. Mistral AI records $1.5 /M tok, 2nd of 9 in the LLM APIs roster. We define that field as pay-as-you-go list price per million output tokens for the named flagship model, including reasoning or thinking tokens where those are billed as output.
• $/M input is your binding constraint. Mistral AI records $0.5 /M tok, 2nd of 9 in the LLM APIs roster. We define that field as pay-as-you-go list price per million input tokens for the named flagship model, uncached, at the standard context tier.
• $/M cache read is your binding constraint. Mistral AI records $0.05 /M tok, 2nd of 8 in the LLM APIs roster.

Reach for something else when any of the following is a requirement rather than a nice-to-have.

• Context window. Mistral AI records 256,000 tokens, 10th of 12 in the LLM APIs roster. Anthropic records 1,000,000 tokens on the same field. If that is a hard requirement rather than a preference, start elsewhere.

• We publish this block because a comparison that only lists what a tool is good at is marketing.
• Every figure above sits on the same page as its source, and the field definitions are on the category tables if you want to check how we measured them.

What we don't know.

• 5 of the 27 fields we track for Mistral AI are still blank: Max output, TTFT p50, Output tok/s, Cache TTL and Notice period.
• Those render as dashes rather than as zeroes or assumptions, because an empty cell and a bad cell are not the same thing and only one of them is honest.
• If you know any of these figures and can point at a document, tell us.