ResetToken Guides / Mistral

Mistral AI · Reset Bible

When does your Mistral quota reset?

Honestly? There is no reset moment to wait for. Mistral's API rate limits are rolling windows — requests per second, tokens per minute, tokens per month — that refill continuously as time passes, with no published reset hour or timezone. The one genuine reset in the Mistral system belongs to the consumer side: the Vibe plan budget (Vibe is the app formerly called Le Chat) rolls over on your billing date. This page explains each window, the usage tiers behind them, and where to see your real numbers.

The short answer: hit a Mistral API rate limit and capacity returns as the rolling window slides forward — seconds for the per-second limit, a minute for the per-minute limit — not at midnight, not on the 1st, not at any fixed clock time. Mistral publishes no reset schedule for these windows, so there is no honest countdown to show. Your Vibe plan budget and its bundled API credits are the exception: they reset when your billing period rolls over.

Rules checked · Oct 17, 2026Applies to La Plateforme API & Vibe plans

01 — THE WINDOWSThree rolling windows, no reset clock

Mistral enforces three rate limits at once on the developer API (La Plateforme / Mistral Studio), measured per model across your entire organization. All three are windows that slide with time — none of them is a bucket that empties and then refills at a scheduled moment.

WindowHow it refillsReset momentApplies to
Requests per second (RPS)
rolling · 1 second
Continuously. Each request stops counting against the limit one second after it was made, so capacity frees up request by request, second by second. There is nothing to “wait out” beyond a moment. None. No clock time, no timezone — the window is always the last second, everywhere at once. Every API organization, per model. The first limit a burst of parallel calls will hit.
Tokens per minute (TPM)
rolling · 60 seconds
Continuously, over a sliding 60-second window counting input and output tokens combined. Tokens age out one minute after they were consumed, so headroom returns steadily rather than all at once. None published. Not aligned to the top of the minute or any timezone. Every API organization, per model. Long prompts and long completions eat this window fastest.
Tokens per month
monthly cap
A monthly token ceiling per model sits on top of the two short windows. Unlike them, it does not visibly “slide” — but Mistral does not publicly document where the monthly boundary falls for an account. Not published. No reset day, hour, or timezone is documented — do not trust any site that names one. Every API organization, per model. Your consumption against it is visible in the Console (see §04).
Vibe plan budget
fixed · billing cycle
The consumer plan's monthly budget is the one allowance that behaves like a classic reset: reach it and, per Mistral's help center, the plan can stop working until the next billing period. Your billing-period rollover — the day your Vibe subscription renews. Personal to your account, like every billing-anchored reset. Vibe Free, Pro, Student and Team subscribers (chat product — not the API).

Why this page has no countdown timer: every other Reset Bible page counts down to a real, documented reset instant. For Mistral's API there isn't one — the windows roll. Putting a timer here would mean inventing a reset time Mistral has never published, so instead this page tells you how the windows actually behave and where to read your live headroom: the ratelimitbysize-* headers returned with every API response.

02 — USAGE TIERSYour limits rise with billed usage — not with top-ups

La Plateforme is pay-as-you-go with no subscription tiers to buy. Instead, your organization sits on a usage tier that controls how high the three windows are set, and the tier climbs automatically with your cumulative billed amount. One rule surprises people: prepaid credits do not move your tier — only invoiced, billed usage does.

TierHow you reach itWhat changes
Free mode (default)Automatic on signup — no payment method requiredLimited limits for testing and prototyping. Enough to evaluate the API, not to run production traffic.
Tier 1Enable pay-as-you-goFull pay-per-token billing unlocks, with the first step up in rate limits.
Tier 2Over €20 / $20 billed in totalHigher RPS, tokens-per-minute and tokens-per-month ceilings.
Tier 3Over €100 / $100 billed in totalHigher ceilings again — thresholds are cumulative, so every invoice counts toward the next tier.
Tier 4Over €500 / $500 billed in totalHigher ceilings again.
Above Tier 4Over €2,000 / $2,000 billed in total, then contact supportThe highest standard ceilings; beyond that, limits are arranged with Mistral support.
  • The ceilings themselves are not public. Mistral does not publish the numeric RPS / TPM / monthly-token values for any tier — they are shown only in the signed-in Admin panel (see §04). Third-party tables quoting exact numbers conflict with each other; treat them as unverified.
  • Credits ≠ tier. Buying prepaid API credits funds your balance, but Mistral states explicitly that adding credits does not raise rate limits. Credits are also applied automatically before your payment method is charged — and credits added after an invoice is issued can't cover that invoice.
  • Free mode is a real mode, not a trial clock. It carries limited limits for prototyping and ends when you enable pay-as-you-go — there is no day-count expiry attached to it in Mistral's documentation.

03 — VIBE & THE PRO CREDIT BUNDLEThe chat app formerly known as Le Chat

In May 2026 Mistral renamed its consumer chat product from Le Chat to Vibe — same accounts, same plans, history carried over. Vibe has its own plan budget, separate from API rate limits, and in 2026 paid Vibe plans started bundling monthly API credits — the bridge between the chat subscription and the developer API.

Vibe Free — $0

Limited messages and web searches, limited coding sessions, basic image generation and editing, 100+ connectors, and up to 5 scheduled tasks. No API credits are listed for the free plan on Mistral's pricing page.

Vibe Pro — $14.99/mo

More messages and web searches (up to 6× free messages, 5× free web searches per the official comparison), faster responses, all-day coding, and $25.50/month in API credits. Price excludes tax.

Student & Team

Verified students get Pro at $5.99/mo with $15/month in API credits. Team runs $24.99/user/mo (excluding taxes) with higher limits and up to 30GB storage per user.

  • The bundled credits reset with the billing cycle. The $25.50 on Pro is a monthly allowance: a fresh bundle arrives each billing period, tied to the same rollover that resets the Vibe plan budget. Mistral has not explicitly documented rollover for these credits, so treat them as use-it-or-lose-it within the period.
  • Credits are money, not headroom. Bundled or prepaid API credits pay for usage; they do not lift the RPS / TPM / monthly-token windows. A funded account can still be rate-limited, and an exhausted credit bundle doesn't change when your windows refill — they were never empty in that sense.
  • Two systems, billed separately. Mistral's help center is explicit: API rate limits are not the same as your Vibe plan budget. Rate limits govern the developer API and Studio; the plan budget governs the Vibe apps — different counters, different resets (rolling windows vs. billing rollover).
  • Prepaid credit expiry: not published. Mistral states no expiry policy for prepaid API credits in its billing documentation. Anyone quoting you a fixed expiry (12 months is a common guess) is extrapolating from other providers.

04 — YOUR NUMBERSWhere to check your own usage & limits

Because Mistral keeps the numeric ceilings behind sign-in, your own console is the only authoritative source. Three official places to look:

Console — organization usage

console.mistral.ai

The La Plateforme console holds your API keys and the “View organization usage” readout, showing monthly API usage — and Vibe quota usage as a percentage — for your organization.

Open the Mistral Console ↗

Admin panel — Limits page

admin.mistral.ai

The signed-in Admin panel's Limits page is where Mistral shows your organization's actual tier and per-model rate-limit ceilings — the numbers that appear in no public documentation.

Open the Admin panel ↗

Help Center — rate limits explained

help.mistral.ai

Mistral's own article on why you're hitting rate limits: the three limit types, the cumulative-billing tier ladder, and the distinction between API rate limits and the Vibe plan budget.

Read the rate-limits article ↗ Plans & API pricing ↗

05 — FAQMistral reset questions, answered straight

When does my Mistral API quota reset?

There is no fixed reset moment to wait for. Mistral's API rate limits are rolling windows — requests per second, tokens per minute, and tokens per month — enforced per model at the organization level. The per-second and per-minute windows refill continuously as time passes; Mistral publishes no reset clock, hour, or timezone for any of them.

Is there a Mistral reset time or timezone I can count down to?

No. Unlike subscription windows that reset at a stated hour, Mistral publishes no fixed reset time or timezone for its API windows, and the exact boundary of the tokens-per-month window is not publicly documented. Any site showing a Mistral API reset countdown has invented it. The only dated reset in the Mistral system is your Vibe plan budget, which rolls over with your billing period.

What are Mistral's three API rate limits?

Three limits are enforced at the same time, per model, across your whole organization: requests per second (RPS), tokens per minute (input plus output combined), and tokens per month. Hitting any one of them throttles you, even if the other two still have room — a burst of parallel calls trips RPS first; long generations trip TPM.

Why can't I find my exact Mistral rate-limit numbers?

Because Mistral does not publish numeric ceilings per tier. Your actual limits are shown only in the signed-in Admin panel, on the Limits page. Every API response also carries ratelimitbysize-* headers, which are the live source of truth for the model you just called. Third-party sites quoting exact per-tier numbers disagree with each other — treat them as unverified.

How do I raise my Mistral API rate limits?

Your usage tier rises automatically with your cumulative billed amount: enabling pay-as-you-go puts you on Tier 1, then higher tiers unlock past €20/$20, €100/$100, €500/$500 and €2,000/$2,000 billed, with custom limits via support beyond that. Note the trap: adding prepaid credits does not raise your rate limits — only billed usage does.

What happened to Le Chat?

Le Chat was renamed Vibe in May 2026. Existing accounts and plans carried over, and chat history is preserved. Vibe is Mistral's consumer chat product; it is a separate system from the developer API (La Plateforme), with its own plan budget — though Vibe plans now bundle monthly API credits.

Does the Vibe Pro plan include API credits?

Yes — $25.50/month in API credits on Vibe Pro ($14.99/month, excluding tax), per Mistral's official pricing page. The student plan ($5.99/month for verified students) bundles $15/month in API credits. These credits are a monthly allowance tied to your billing cycle — and they are separate from the API's rate limits, which still apply to every call the credits pay for.

When does my Vibe plan budget reset?

At your subscription's billing-period rollover. Mistral's help center states that when the plan budget is reached, the plan can stop working until the next billing period. So the reset day is your billing date — the day your Vibe subscription renews — not a universal calendar date shared with other subscribers.

Do unused Mistral API credits expire or roll over?

Mistral does not publish an expiry policy for prepaid API credits, so no expiry date can be stated honestly. The plan-bundled monthly API credits (such as Pro's $25.50) are a monthly allowance by design, but Mistral has not explicitly documented whether they roll over — assume they are use-it-or-lose-it within each billing period and check your console balance.

Are Vibe plan limits the same as API rate limits?

No. Mistral's own help center separates them: API rate limits (RPS, tokens per minute, tokens per month) apply to the developer API and Mistral Studio, while the Vibe plan budget governs the consumer chat product. The two are billed separately — the bridge between them is the monthly API-credit bundle included in paid Vibe plans.