Mistral Medium 3.5 → Mistral Large 4: cost and quality on real tasks
Mistral shuts down mistral-medium-3-5 on 2027-04-30 and names Mistral Large 4 as its replacement (Mistral's deprecations page). After that date, requests to Mistral Medium 3.5 fail.
On the six tasks below, Mistral Large 4 costs 54–69% less per request at list prices.
Price per million tokens
Standard list prices, checked 2026-11-12. Prices change often, so check the source before you rely on them.
| Model | Input | Cached input | Output | Source |
|---|---|---|---|---|
| Mistral Medium 3.5 (current)mistralai/mistral-medium-3-5 | $1.50 | Not listed | $7.50 | OpenRouter listing |
| Mistral Large 4 (new)mistralai/mistral-large-4-0 | $0.70 | $0.07 | $2.10 | OpenRouter listing |
- Note on Mistral Medium 3.5: Mistral's own endpoint on OpenRouter. Deprecated 2026-10-31, retires 2027-04-30 (replacement Mistral Large 4).
- Note on Mistral Large 4: Price of Mistral's own endpoint on OpenRouter. mistral.ai/pricing no longer shows a per-model API table (its FAQ example says 'Mistral Large $0.5/$1.5', which matches Large 3). Some third-party sites report a $1.36/$4.18 list price; unverified.
Cost per 1,000 results on six common tasks
What 1,000 requests cost at list price, for six task shapes from different kinds of app. The token counts are illustrative, not measured: your prompts may be longer or shorter, and the switch test below shows your real ones.
| Task, and what goes in and out | Tokens in / out | Mistral Medium 3.5 | Mistral Large 4 | Change | Break-even: Mistral Large 4 costs no more per result that works if it works first time at least |
|---|---|---|---|---|---|
| Support bot answerInstructions, three help-centre passages and the customer's question in; a short answer out. | 3,000 / 300 | $6.75 | $2.73 | 60% less | 41% as often as Mistral Medium 3.5 |
| Invoice field extractionInstructions and one invoice's text in; about 15 fields as JSON out. | 2,500 / 250 | $5.63 | $2.28 | 60% less | 41% as often as Mistral Medium 3.5 |
| Lead scoringScoring rules and one lead's form answers in; a score and a one-line reason out. | 1,200 / 100 | $2.55 | $1.05 | 59% less | 42% as often as Mistral Medium 3.5 |
| Email reply draftInstructions and the email thread so far in; a reply of about 300 words out. | 1,500 / 400 | $5.25 | $1.89 | 64% less | 36% as often as Mistral Medium 3.5 |
| TranslationInstructions and about 500 words of text in; the same text in another language out. | 900 / 800 | $7.35 | $2.31 | 69% less | 32% as often as Mistral Medium 3.5 |
| Search re-rankingA search query and 20 result snippets in; the ids of the best five, in order, out. | 4,000 / 50 | $6.38 | $2.91 | 54% less | 46% as often as Mistral Medium 3.5 |
Real cost per result: when a switch really saves
Real cost per result is what you actually pay for each result that worked, counting the failed tries. A result that fails gets tried again, and you pay for both tries.
So the price per request is only half the answer. Mistral Large 4 stays cheaper per result that works as long as it works first time at least 32% as often as Mistral Medium 3.5 on translation, and 46% as often on search re-ranking. That's the new cost divided by the current cost, shown for each task in the last column above.
Example, with a made-up rate
Say 8 in 10 of Mistral Medium 3.5's support bot answers work first time. You pay $6.75 for 1,000 tries and get 800 that work, a real cost of $8.44 per 1,000 support bot answers that work. Mistral Large 4 costs $2.73 for 1,000 tries, so it matches that real cost if 33% of its support bot answers work first time, and costs less if more do.
Quality: test it on your own requests
This page can't tell you whether Mistral Large 4 handles your requests as well as Mistral Medium 3.5, and we don't publish quality scores. Quality depends on your prompts and your inputs: a model that answers short English questions well can still fail on long invoices, odd layouts or other languages.
The free model switch test runs your prompt and up to 20 of your real inputs on both models with your own key, through each provider's own API or OpenRouter. You see the answers side by side with the tokens and cost of each, and mark which ones worked.
Test both models on your promptsWhat to compare
- Format: does JSON still parse with the same field names, for extraction and scoring?
- Length and tone: are support answers and email drafts as short and polite as before?
- Refusals: does Mistral Large 4 decline requests that Mistral Medium 3.5 handled?
- Hard inputs: long documents, other languages and messy formatting, not just the easy cases.
Questions
Is Mistral Large 4 cheaper than Mistral Medium 3.5?
Per request, yes: it costs 54–69% less on six common tasks. Mistral Medium 3.5 lists at $1.50 per million input tokens and $7.50 per million output tokens; Mistral Large 4 lists at $0.70 and $2.10 (Mistral Medium 3.5: price checked 12 nov 2026; unchanged since we started tracking it; Mistral Large 4: price checked 12 nov 2026; unchanged since we started tracking it). Per result that works, it depends on how often each model works first time on your inputs: Mistral Large 4 stays cheaper per result that works as long as it works first time at least 32% as often as Mistral Medium 3.5 on translation, and 46% as often on search re-ranking.
When does Mistral Medium 3.5 stop working?
Mistral shuts down mistral-medium-3-5 on 2027-04-30, according to its deprecations page. Requests that name it will fail after that, so test the replacement on your real requests and switch before then.
Will Mistral Large 4 work as well as Mistral Medium 3.5 for my app?
Nobody can tell you that from a price table or a general benchmark, because it depends on your prompts and inputs. Run 10 to 20 of your real requests through both models with the free model switch test, using your own provider key, and compare the answers side by side before you switch any users.