How Much Is Grok 4? | Current Plans And API Rates

“Grok 4” can mean two different things: using Grok inside an app, or calling Grok through an API for a product you’re building. The price depends on which path you pick, how many messages you send, and whether you want the bigger “Heavy” tier.

This breakdown keeps it practical. You’ll see the real monthly numbers people end up paying, where the extra dollars come from (tax, app store fees, region), and a quick way to estimate API spend without guessing.

What You’re Paying For With Grok 4 Access

At a high level, Grok 4 pricing has three buckets:

• App subscription (SuperGrok): pay a flat monthly fee and use Grok on grok.com or the Grok apps, with higher usage limits than free access.
• X subscription (Premium+): pay a flat monthly fee that bundles Grok with X features.
• API usage (xAI API): pay per token for requests and outputs, like a utility bill tied to how much your software uses.

Each bucket fits a different kind of user. If you chat a lot and want predictable billing, subscriptions feel clean. If you’re building a tool, a bot, or a workflow for a team, API pricing usually maps better to actual usage.

How Much Is Grok 4? In Real Terms

Most people want a straight number. Here’s the cleanest way to read it:

• SuperGrok: commonly listed at $30/month (or a discounted annual option) for access to Grok via grok.com and the Grok apps.
• SuperGrok Heavy: commonly listed at $300/month for the heavier tier tied to Grok 4 Heavy.
• X Premium+: “starts at” $40/month on web pricing, with regional pricing and taxes layered on top.

Two quick caveats save headaches:

• Region and tax: many pricing pages show “starts at” language. Your checkout total can shift by country and tax rules.
• Web vs in-app: Apple/Google billing can differ from web billing. If you’re price-sensitive, check the web checkout before subscribing in an app.

Choosing Between SuperGrok And X Premium+

If you only care about the assistant, SuperGrok is the simpler bill. If you already use X every day and want the platform perks, Premium+ can feel like one subscription that covers both.

SuperGrok Is The Straight “Just Give Me Grok” Plan

SuperGrok is positioned as a direct route to Grok with higher limits and feature access through grok.com and the standalone apps. If your main habit is chatting, drafting, coding, and generating images inside Grok itself, this is the least cluttered choice.

If you’re trying to keep subscriptions lean, SuperGrok also makes it easier to track what you’re paying for. You aren’t paying for extra platform features you might never touch.

X Premium+ Makes Sense If You Want The X Bundle

X Premium+ is a bundle. Grok access comes with X-side features like reduced ads, posting tools, and other plan benefits. X’s help docs describe Premium+ as starting at $40/month on web, with regional pricing differences.

If you’re already paying for Premium+ for X features, Grok can feel “included.” If you’re subscribing only to get Grok, compare that same monthly number to SuperGrok first.

What “Free Grok” Usually Means

Free access tends to be limited by daily usage caps, slower lanes during peak times, or feature restrictions. It can be fine for occasional questions. It gets frustrating fast if you’re doing longer writing, multi-step coding, or repeated iterations.

A simple test: if you hit limits more than a couple of times a week, you’re already spending time paying a “friction tax.” That’s usually the moment a flat subscription starts to feel reasonable.

Grok 4 Pricing By Access Route And Typical User

Below is a simple way to pick the right payment model without overthinking it. This table is intentionally broad so you can map your own situation to a plan.

Access Route	Typical Price Point	Best Fit
SuperGrok (via grok.com)	$30/month (often with annual option)	Heavy chat users who want one bill and don’t need X perks
SuperGrok Heavy (via grok.com)	$300/month (often with annual option)	Power users running hard reasoning tasks and long sessions
X Premium+ (web pricing)	Starts at $40/month	People who want Grok plus the Premium+ bundle inside X
xAI API (pay-as-you-go)	From $2/1M input tokens + $6/1M output tokens (model-dependent)	Builders shipping apps, agents, and automations
xAI API (fast variants)	Lower token rates on “fast” models	Latency-sensitive tools and high-volume endpoints
Hybrid: subscription + API	Two bills	Creators who chat personally, plus a product that calls the API
Team path (varies by product)	Seat-based or contract pricing	Teams that need admin controls and billing centralization

The API Route: When Paying Per Token Is Cheaper

If you’re building anything that calls Grok from code, subscription pricing isn’t the right measuring stick. You need token pricing. The xAI API lists model rates per million tokens, split into input and output.

Here’s the mental model that works:

• Input tokens: what you send (your prompt, system instructions, tool results you include, chat history).
• Output tokens: what Grok returns (the answer, structured JSON, tool plans).

Two usage patterns swing your bill more than people expect:

• Long chat history: if you keep sending a giant thread each request, input tokens climb quickly.
• Big outputs: long drafts, multi-file code, or verbose reasoning can push output tokens higher than input.

If you want the official rates and model names, use the xAI pricing page for the API: xAI API models and pricing.

API Cost Estimation Without Guesswork

You don’t need perfect token math to estimate spend. Start with a rough “per request” shape, multiply by daily volume, then sanity check after a day of logs.

Step 1: Pick A Typical Request Shape

Most apps fall into one of these shapes:

• Short Q&A: small input, small output (think: tooltips, quick answers).
• Workflow: medium input, medium output (think: summarization, parsing, classification).
• Long-form generation: medium input, large output (think: drafting, code scaffolds).

Step 2: Multiply By Volume

If your endpoint serves 1,000 requests/day, a few cents per call becomes real money. If you serve 50 requests/day, even the flagship model can be manageable.

Step 3: Track Two Numbers In Your Logs

Log input tokens and output tokens for each call. After a day, you’ll know your true median and your worst-case spikes. That’s enough to set budgets and guardrails.

Sample Monthly API Scenarios Using Listed Rates

This table gives you quick math using the published per-million token rates. It’s meant to be directional, then you refine it with your own logs.

Scenario	Monthly Token Use	Estimated Cost At Listed Rates
Prototype app (light)	10M input + 10M output	(10×$2) + (10×$6) = $80/month
Internal tool (moderate)	50M input + 25M output	(50×$2) + (25×$6) = $250/month
Content pipeline (output-heavy)	30M input + 60M output	(30×$2) + (60×$6) = $420/month
High-volume endpoint (fast model)	200M input + 100M output	Rates vary by model; fast tiers can cut this sharply
Agentic workflow (history-heavy)	120M input + 40M output	Input dominates; trimming history usually drops spend

Hidden Cost Drivers People Miss

Regional Pricing And Taxes

Subscription totals can shift based on your country, local taxes, and payment method. “Starts at” pricing is a sign your checkout may differ. If you’re comparing plans, compare the final checkout totals, not just the headline number.

App Store Billing

Buying through an app can carry different pricing than buying on web. If you want the lowest straightforward price, check the web plan page first: Grok plans on grok.com.

API Token Waste From Chat History

If you send the full conversation each turn, your app pays repeatedly for old text. A clean fix is to summarize older turns and keep only what the model needs for the next step. That single change can drop input token totals a lot.

Output Creep

Longer answers feel nice, then output tokens quietly become the main line item. Put a sensible max output length in place, then build a “continue” button for the rare times a user truly needs more.

Which Option Fits Most People

If You Just Want Grok On Your Screen

Start with SuperGrok if you want a clean monthly bill and you don’t care about X extras. If you already want the Premium+ bundle on X, Premium+ can be the simpler choice since it covers both.

If You’re Building A Product Or Tool

Start with the API and pick the model tier that matches your latency and budget needs. You can also run a fast model for everyday calls and reserve the flagship tier for tougher requests.

If You’re On The Fence

Do a one-week test. Track how often you hit free limits and how many minutes you lose restarting prompts. If that friction shows up again and again, you’ll know a subscription is worth it for you. If it doesn’t, stay free and spend the money elsewhere.