Quick answer
Grok Free costs $0, but xAI does not publish a universal number of free chat messages or a guaranteed reset interval on its pricing page. It describes Free use as limited; xAI’s consumer FAQ says Free Chat and Voice limits are separate from the paid weekly usage pool and reset on their own schedule. Check the notice shown in your account for the limit that applies to you. (xAI pricing; Grok FAQ)
One third-party author reported seeing a wait notice after about 10 messages on grok.com during a period estimated at roughly two hours. That is an account-specific observation—not a published xAI quota, a timed confirmation of a reset rule, or a forecast for your account. The author said their notes did not establish the same count on X. (Nexoda Tech)
For production development, the consumer app’s free limit is not your API allowance. xAI publishes separate API prices and rate limits. Its API pricing documentation lists Grok 4.7 token rates; API spend depends on the model and tokens used. (API pricing)
What is—and isn’t—published about Grok Free limits
| Question | What the sources establish |
|---|---|
| Does Free cost anything? | $0/month on xAI’s pricing page. (xAI pricing) |
| Is there a published universal free message count? | xAI describes Free as limited, including limited chat requests, but does not give a universal message count on the pricing page. (xAI pricing) |
| Is there a published universal free reset time? | The consumer FAQ says Free Chat and Voice limits reset on their own schedule, separately from the paid weekly pool; it does not specify a universal interval in the cited guidance. (Grok FAQ) |
| Does “about 10 in roughly two hours” apply to everyone? | No. It is one author’s approximate grok.com observation, not an official or guaranteed quota. (Nexoda Tech) |
So, xAI does not give a single public message count to plan around; check the current limit notice for your account. Do not assume a third-party test or a notice on another surface describes your allowance on grok.com, X, or the mobile apps.
Free versus paid consumer plans
The reported monthly prices for The X Subscription are below. They are monthly USD list prices; check xAI’s live pricing page for current terms before purchasing. (xAI pricing)
| Plan | Listed price per month (USD) | Usage description |
|---|---|---|
| Free | $0 | Limited AI use, including limited chat and agent requests |
| Plus | $8 | Base usage pool |
| Premium | $30 | 4× the Plus usage amount |
| Super | $100 | 15× the Plus usage amount; priority model access at peak times |
| Ultra | $200 | 50× the Plus usage amount |
If you only need occasional chats, Free avoids a subscription charge, but its exact allowance is not stated as a universal number. A paid consumer plan may provide more included usage; compare the current plan details with your needs. A consumer subscription is not the per-token price for production API requests.
Consumer Grok and the production API are different budgets
The Grok website and apps are consumer products. The developer API has separate model pricing and rate limits. A free consumer account—or a paid consumer plan—does not establish how much API traffic an application can send or what those requests will cost. xAI describes API throughput limits in two dimensions: requests per second (RPS) and tokens per minute (TPM), with limits set per model and team. Check the current API rate-limit documentation for the applicable limits. (API rate limits)
For a production API reference, xAI’s Grok 4.7 model page lists a 500,000-token context window, configurable reasoning, and function calling. These are documented model capabilities, not evidence of accuracy or suitability for a particular application. (Grok 4.7 model documentation)
For short-context Grok 4.7 text API requests, the listed rates are:
| API item | Price per 1 million tokens (USD) |
|---|---|
| Input | $2.00 |
| Cached input | $0.50 |
| Output | $6.00 |
| Long-context input | $4.00 |
| Long-context cached input | $1.00 |
| Long-context output | $12.00 |
The pricing documentation identifies a 200,000-token threshold for long-context rates. Use the applicable rate table for your request rather than assuming the short-context prices apply. Requests sent to the U.S. regional endpoint are listed at 1.1× global token rates. Confirm current rates and endpoint terms in xAI’s API pricing documentation before budgeting. (API pricing)
A reproducible API cost example
Suppose an application sends 1,000 requests, each with 2,000 uncached input tokens and 500 output tokens. Assume every request falls under the short-context rates and uses no other billable service:
- Input: 1,000 × 2,000 = 2,000,000 tokens; 2 × $2 = $4.
- Output: 1,000 × 500 = 500,000 tokens; 0.5 × $6 = $3.
- Estimated model-token total: $7 USD.
This is an illustration, not a quote. Actual spend depends on the tokens sent—including conversation history—and the applicable model, context, cached-input, and endpoint rates. Check whether any separately selected API service has its own charge.
How to check before relying on Grok
- For Free consumer access: Open Grok on the surface you plan to use and follow the limit notice shown for your account. Do not treat the reported 10-message observation as a guaranteed quota.
- For a consumer subscription: Compare current monthly prices and plan details on xAI’s pricing page before purchasing. (xAI pricing)
- For an API deployment: Confirm the model identifier, per-token rates, endpoint terms, and your team’s per-model RPS and TPM limits in xAI’s current API documentation. Estimate cost from expected input and output tokens, then validate with a representative pilot. (API pricing; API rate limits)
Bottom line: Grok Free is $0 with limited use, but the published consumer plan does not promise a universal message count or reset interval. Treat “about 10 in roughly two hours” as one third-party report, not a rule. Budget production API use separately using the relevant model prices and rate limits.