Is Grok Free? Grok Pricing (2026): Where the Free Plan Stops

Yes. Grok Free costs $0 and now includes Grok Build, image generation, Voice, connectors, and limited real-time web and X search, but xAI still publishes no stable message allowance. The paid ladder is Lite at $8/month, SuperGrok at $30, Plus at $100, and Heavy at $300.

Grok pricing at a glance

Grok is free to start, and most casual users should keep it that way. The paid ladder buys more capacity and named features, but xAI's public pages do not convert that capacity into a guaranteed number of prompts, images, videos, or coding tasks.

Every price, tier, and limit below was verified against xAI's live pricing page, grok.com, and xAI's consumer and developer documentation on 24 September 2026.

Since September 21, 2026, Grok 4.7 has been xAI's frontier model for coding and knowledge work, available in Grok Build, the API, Cursor, and model gateways. The consumer pricing comparison still labels its plan model as Grok 4.6, while Grok 4.7 is the default in Build; the faster 4.7 variant is excluded from Build's free tier. That distinction matters because a subscription tier does not promise 4.7 on every Grok surface.

Business now has a public $30 per-user monthly price, while Enterprise still routes buyers to sales. The short verdict remains blunt: stay Free while a reset is merely inconvenient, buy Lite when interruptions recur, choose SuperGrok when Bot and higher limits become weekly tools, add Plus for sustained high usage and 1080p video, and reserve Heavy for work that can defend $300.

What Grok Free actually gives you

Grok Free is a continuing $0 plan, not a timed trial. xAI's current comparison includes Grok 4.6, image generation, Grok Build, Voice, connectors, and web, iOS, and Android apps. Real-time web and X search are marked limited, while video generation, Expert, Grok Bot, and priority support sit above Free.

The limit is real, but the number is not public. xAI documents separate Free limits for Chat and Voice, each on its own reset schedule, yet it does not promise a stable count. A prompt count observed on one account is not a price-sheet guarantee because model load, feature, platform, and product changes can move the effective allowance.

That uncertainty is the Free plan's main cost. A senior operator can spend $0 in cash and still lose a useful work session when a cap arrives with no published capacity model. A casual user who asks a few questions, checks X context, or uses Voice occasionally may never feel it.

The upgrade trigger should be an event, not curiosity. Write down each time Free blocks a useful outcome: a finished brief, a research decision, a coding task, or a media deliverable. Repeated annoyance is still not a business case unless the interruption delays work worth more than the next plan.

Free also provides a clean way to judge product fit. Grok's differentiator is its direct access to current web and X context, now joined by image generation and Build. If those surfaces do not improve your work while the plan costs $0, paying for more of them will not fix the mismatch.

What each paid Grok plan buys

Paid Grok plans buy a larger shared weekly allowance, not unlimited usage. Since June 2026, xAI has been moving paid customers from separate daily caps for Chat, Imagine, Voice, and Build to one weekly usage pool that can be spent across those products.

The pool is shown as a percentage in Settings → Usage, together with a product breakdown and reset time. A text chat uses less compute than high-quality video or a long coding task, so the percentage does not map to a fixed number of actions. That flexibility is useful, but the missing conversion rate prevents exact plan-to-plan capacity comparisons.

SuperGrok Lite: the $8 interruption fix

SuperGrok Lite is the sensible first purchase for a light user who likes Grok but repeatedly hits Free. At $8/month, it adds video generation, Expert, and full real-time web and X search to the Free capabilities. It includes Build and image generation, but not Grok Bot or priority support.

Lite's weakness is visibility. xAI does not expose a stable included-action count, and its public pricing page quotes only the monthly amount. Buy it monthly as a controlled experiment. If the same cap still interrupts work, the result is useful evidence for SuperGrok rather than a reason to stack random top-ups forever.

SuperGrok: the $30 Bot plan

SuperGrok is the default paid plan for someone using several Grok surfaces every week. It costs $30/month and adds Grok Bot plus higher rate limits across all features. The consumer pricing page still lists Grok 4.6, Expert, connectors, and image and video generation; Grok 4.7 belongs to the separate Build and API availability described above.

xAI's public pricing page no longer publishes a consumer annual dollar amount. The FAQ acknowledges yearly Heavy subscriptions, so an annual option may still appear in an account checkout, but any savings calculation should use the amount shown at purchase rather than the old public figures.

SuperGrok Plus: the $100 high-usage plan

SuperGrok Plus fills the old gap between SuperGrok and Heavy. At $100/month, it adds 1080p video, significantly higher usage across Chat, Imagine, Voice, and Build, faster replies, priority access at peak times, and early access to new features.

Plus is the rational step when a $30 plan repeatedly pauses paid work but Heavy is still impossible to defend. It costs $70 more than SuperGrok and $200 less than Heavy, so it creates a much cleaner upgrade decision than jumping straight from $30 to $300.

SuperGrok Heavy: a $300 capacity purchase

SuperGrok Heavy is for measured exhaustion, not status. The plan costs $300/month and adds xAI's highest usage at its fastest speed, capacity for difficult problems, dedicated support, and early access.

Heavy costs ten times SuperGrok and three times Plus. xAI does not publish a matching action guarantee, so the tier cannot be justified from the card alone. A funded founder running repeatable research or media workflows can consider it after Plus repeatedly stops paid work. A mid-market CTO should require usage records, a named workload, and a monthly value estimate before approving it.

The break-even and cost per outcome

The current monthly ladder has four upgrade gaps: $8 from Free to Lite, $22 from Lite to SuperGrok, $70 from SuperGrok to Plus, and $200 from Plus to Heavy. Those gaps, not the headline prices alone, are the amounts an interruption must earn back.

xAI's public page does not quote consumer annual prices. Do not carry forward previously published annual figures. If an annual offer appears in your logged-in checkout, compare that exact total with the number of months you expect to keep the plan.

Top-ups create a second break-even. Lite plus $22 of Extra Usage Credits reaches SuperGrok's $30 sticker price. SuperGrok plus $70 reaches Plus at $100, and Plus plus $200 reaches Heavy at $300. These are spend break-evens, not capacity break-evens, because xAI does not publish how much included usage each upgrade adds.

Cost per outcome is still useful if you define the outcome yourself. At 100 completed outcomes in a month, Lite allocates $0.08 to each, SuperGrok $0.30, Plus $1.00, and Heavy $3.00. At only 10 useful outcomes, those figures become $0.80, $3.00, $10.00, and $30.00. An outcome is a finished decision, approved document, working code task, or shipped media asset, not another prompt.

That framing changes the purchase. A solo technical builder who completes 100 useful coding and research outcomes with SuperGrok is allocating $0.30 to each. A Plus subscriber who produces only 10 useful outcomes is allocating $10.00 to each. The same plan can be cheap or wasteful without its list price changing.

1. 1

   Stay on Free long enough to observe friction

   Record blocked outcomes for one normal week. Do not count curiosity, only work you would have completed.

2. 2

   Buy the next tier monthly

   Choose Lite first unless you already need SuperGrok's named features. Monthly billing preserves the option to leave.

3. 3

   Check Settings → Usage before topping up

   Note which product consumed the weekly pool. Video, Voice, Build, and Chat do not draw equal compute.

4. 4

   Recheck checkout before an annual commitment

   The public pricing page currently quotes monthly consumer amounts only. If your account shows an annual offer, calculate its break-even from that live checkout total.

Hidden costs and billing catches

The largest hidden cost is open-ended top-up spend. Extra Usage Credits start at $5, can currently be purchased only on the web, and are consumed after the included weekly allowance. xAI says their cost per action is higher than the effective rate inside a plan.

Credits expire one year after purchase. Auto Top Up can refill them when the balance falls, with an amount and monthly cap chosen by the account owner. Set that cap before the first purchase. A $30 subscription with an unreviewed refill rule is no longer a fixed $30 product.

Any annual offer has a separate downside. xAI's consumer terms say subscriptions renew automatically and payments already made are generally non-refundable unless law requires otherwise. The company promises reasonable notice before a price increase applies at renewal, not a fixed 30-day notice period, and that does not make an unused annual plan liquid.

Refunds depend on the purchase channel. xAI reviews requests for web and Google Play purchases, and says approved refunds usually reach the original payment method within 5 to 10 business days. Apple handles App Store refunds. A review process is not a refund promise, so the controlling assumption should remain non-refundable.

Account identity can also create accidental double billing. A subscription belongs to the Apple, Google, email, or X login used at purchase. Signing into another identity can make a paid plan appear missing. Link the X account and confirm the active login in Settings before paying again.

Mobile prices, local taxes, currency conversion, and payment-method fees can change the checkout total. The prices here are current US web prices. Compare the final amount on the exact platform where you intend to manage the subscription.

xAI publishes no standing student plan or student discount on its pricing page, consumer FAQ, or consumer terms. Students can use Free, but there is no verified reduced SuperGrok rate to budget around.

Grok app, X Premium, and API are separate meters

Grok has three buying routes that should not be blended: the standalone Grok subscription, an X Premium bundle, and the xAI API. Paying on one route does not create a general pool of dollars for the others.

X Premium is an X purchase with Grok benefits

X Premium costs $8/month or $84/year on the US web checkout and includes increased Grok limits. Premium+ costs $40/month or $395/year and now includes SuperGrok access, Grok Bot, Imagine, and Voice alongside its X features. Basic costs $3/month or $32/year, but its public feature list does not promise increased Grok limits.

Lite and X Premium now cost the same $8/month, so choose Premium when the checkmark, reduced ads, creator tools, and X features matter alongside higher Grok limits. Choose standalone SuperGrok when the AI product is the purchase. At $30, SuperGrok is $10 below Premium+, while Premium+ earns its extra cost by bundling SuperGrok and Grok Bot with the X subscription.

The API charges for usage separately

The xAI API is the right route for software, automation, and volume you want to meter directly. A consumer subscription is not an API credit balance: API-key calls use xAI's usage-based API meter, and API credits are non-refundable.

For Grok 4.7 below 200,000 prompt tokens, xAI charges $2 per million fresh input tokens, $0.50 per million cached input tokens, and $6 per million output tokens. A normalized exchange with 1,000 fresh input and 1,000 output tokens costs $0.008. At that shape, $30 buys 3,750 exchanges before tools, retries, storage, or priority processing.

Reach 200,000 prompt tokens and the long-context rates apply to every token in the request: $4 input, $1 cached input, and $12 output per million. Priority Processing doubles standard token rates. Web search and code execution each add $5 per 1,000 calls; X search now costs $5 per 1,000 posts fetched and $10 per 1,000 profiles fetched.

Media has clearer outcome pricing. A 1K Low image from grok-imagine-image-2.0 costs $0.04 to output, plus $0.01 when an image is supplied as input. On output price alone, $30 equals 750 images. A 30-second 720p video on grok-imagine-video-1.5 costs $4.20, while 1080p costs $7.50, plus a possible $0.01 image input. Seven 720p videos cost $29.40; eight cost $33.60.

API storage can keep billing after generation. File storage is $0.025 per GiB per day, collection storage is $0.10 per GiB per day, and downloads are $0.20 per GiB. The Grok 4.5 review works through the model's broader token and context economics.

The decision rule is practical. Use the app for interactive work spanning chat, voice, media, and coding. Use the API when you need programmatic control and a countable bill. Do not buy a consumer plan expecting it to fund production calls.

Grok versus ChatGPT and Claude on price

ChatGPT matches Grok at the entry paid tier and remains cheaper at the mainstream tier. Its current ladder is $0 for Free, $8 for Go, $20 for Plus, and from $100 for Pro.

Grok Lite and ChatGPT Go both cost $8. SuperGrok costs $10 more than ChatGPT Plus, a 50% premium. SuperGrok Plus matches ChatGPT Pro's $100 starting price, while Heavy at $300 is three times that starting price. Those differences buy different products, so the right comparison is the workflow you need, not the longest feature list. The full current ladder is in the ChatGPT pricing guide.

Claude is tied with ChatGPT at the mainstream monthly tier. Claude Free costs $0, Claude Pro costs $20 month to month or $200 paid up front annually, and Claude Max starts at $100/month.

Annual Claude Pro normalizes to $16.67 per month from a $200 upfront payment. SuperGrok is $10 more than Claude Pro on monthly billing, and xAI's public page does not currently quote a consumer annual amount that would narrow the gap. The Claude Free and pricing analysis explains where its own variable limits stop being economical.

Choose Grok when current X context, Grok's media stack, or its product-specific workflow earns the premium. Choose ChatGPT or Claude when their work surfaces fit better and Grok's X advantage does not affect the outcome. A buyer who cannot name the Grok-specific value should not pay 50% more for the mainstream plan.

Which Grok plan should you choose?

Free is the right plan for the casual researcher, occasional voice user, or anyone still deciding whether Grok's X context is useful. Waiting through a cap costs less than a recurring charge when no deadline is attached.

Lite is the right first purchase for a solo user whose Free access interrupts useful work but who does not need Grok Bot or priority support. Its $8 price keeps the experiment small while adding video, Expert, and full real-time search.

SuperGrok is the strongest default for a senior operator or solo builder using Chat, Build, Voice, and Imagine through the week. The $30 price needs a Grok-specific reason because ChatGPT Plus and Claude Pro each cost $20.

Plus is the step for a user who can show that SuperGrok's weekly pool is stopping paid work. At $100, it adds substantially higher usage, faster replies, 1080p video, peak-time priority, and early features without Heavy's $300 commitment.

Heavy belongs to a funded founder or production user with measured high-volume work. A mid-market CTO should skip it for ordinary employee chat: variable shared usage, no public action allowance, and a $300 price make capacity planning too opaque without account data.

The final rule is simple: Free until interruption has a cost, Lite until repeated top-ups approach $22, SuperGrok until measured overages approach $70, Plus until measured overages approach $200, and Heavy only when the workload can defend $300 with completed outcomes.

Frequently asked questions

Is Grok fully free?

Grok has a continuing $0 plan. It includes image generation, Build, Voice, connectors, and limited real-time web and X search, but xAI does not publish a stable numeric allowance.

Does free Grok have a limit?

Yes. xAI documents separate Free limits for Chat and Voice but does not promise a message count. Treat exact limits reported elsewhere as observations, not guaranteed plan terms.

How much does Grok cost?

The consumer ladder is Free at $0, SuperGrok Lite at $8/month, SuperGrok at $30/month, SuperGrok Plus at $100/month, and SuperGrok Heavy at $300/month. Business is $30 per user each month, while Enterprise uses custom pricing. xAI's public pricing page does not currently quote consumer annual amounts.

How can I use Grok without paying?

Sign in at grok.com or use the iOS or Android app and remain within Free limits. You do not need a paid X subscription to start using the standalone Free plan.

Is Grok free on Android and iPhone?

Yes. Grok's $0 plan is available on the web, iOS, and Android. If you later subscribe, manage cancellation and refunds through the channel used for purchase.

Does Grok offer a student discount?

xAI lists no standing student plan or student discount on its current pricing page, consumer FAQ, or consumer terms. Students can use the standard Free plan.

Can I get a refund for Grok or SuperGrok?

xAI reviews web and Google Play requests, while Apple handles App Store purchases. Approved xAI refunds normally arrive within 5 to 10 business days, but the consumer terms say completed payments are generally non-refundable unless law requires otherwise.

Did Grok pricing change in 2026?

Yes. The current pricing page lists SuperGrok Lite at $8 per month, adds SuperGrok Plus at $100, and publishes Business at $30 per user each month. Grok 4.7 is also available through Build and the API, although the consumer comparison still labels its plan model Grok 4.6. Paid plans use one weekly pool across Chat, Imagine, Voice, and Build.

Is the Grok API included with SuperGrok?

No. A consumer subscription is not an API credit balance. API-key use is metered by tokens, media outputs, tools, and storage, and API credits are non-refundable.

Want a faster way to match AI spend to the job each tool should do? Get the AI Tools Map for Business Owners.