The short answer
Grok has a Free plan listed at $0/month for limited use. xAI’s consumer documentation says you can use Grok on grok.com and in its iOS and Android apps, and that Grok is free to start. The cited documentation does not establish whether sign-in is required on each channel. It also does not establish whether the same Free-plan terms apply to Grok inside the X app. (x.ai)
The consumer Free plan and the production API are separate budget questions. xAI lists API charges by token. For Grok 4.7, short-context rates are $2 per million input tokens and $6 per million output tokens. Requests reaching the long-context threshold use higher rates. (docs.x.ai)
What does Grok’s free tier include?
The official plan listing shows Free at $0/month and labels usage as limited. xAI’s Grok overview lists the web app and iOS and Android apps as access channels, and says Grok is free to start. The cited material does not establish whether sign-in is required on each channel, a numeric chat allowance, a reset interval, or which model Free users receive. It also does not establish that the same Free-plan terms apply inside X. (x.ai)
Use Free for manual chats if its limited access meets your needs. If you need application-integrated or other programmatic requests, budget for API usage separately; a $0 consumer plan is not an API price. (x.ai)
Is the production API free?
Do not treat production API calls as free. xAI publishes token-based API prices separately from its consumer plan. The rates below are in USD per 1 million tokens. (docs.x.ai)
Grok 4.7 API text rates
| Billing item | Short context | Long context |
|---|---|---|
| Input tokens | $2.00 | $4.00 |
| Cached input tokens | $0.50 | $1.00 |
| Output tokens | $6.00 | $12.00 |
xAI defines the long-context threshold as 200,000 prompt tokens or more. Once a prompt reaches that threshold, the long-context rates apply to all tokens in that request, not just the tokens beyond 200,000. Grok 4.7’s listed context window is 500,000 tokens. (docs.x.ai)
Server-side tools can add costs beyond token usage: xAI says tool requests are priced using both token usage and server-side tool invocations. Check the listed price for each tool you use. (docs.x.ai)
Example: estimate a small text workload
Suppose your requests total 1 million input tokens and 200,000 output tokens, with every prompt below 200,000 tokens. At the short-context rates, excluding cached input, tools, and other charges:
- Input: 1 × $2.00 = $2.00
- Output: 0.2 × $6.00 = $1.20
- Estimated total: $3.20
This is an arithmetic example, not a bill prediction. Actual cost depends on the tokens each request consumes, caching, whether a prompt reaches the long-context threshold, and any tool invocations. (docs.x.ai)
Free chat or API: which should you choose?
- Choose the Free plan for occasional, hands-on use when limited access is enough.
- Budget for the API when your application or workflow needs programmatic model requests. Estimate input and output tokens separately and include any applicable long-context or tool charges.
- Keep the budgets separate. The consumer Free plan is not an API rate. Check the current plan and developer-pricing pages before committing to a forecast.
For repeatable development tests, saved responses or mocks can reduce how often you need fresh model output. The amount saved depends on your workflow; don’t assume a fixed saving.
Bottom line
Grok’s consumer Free plan is listed at $0/month and offers limited use. xAI documents free access on grok.com and in its iOS and Android apps; the cited material does not establish whether sign-in is required on each channel, whether the same plan applies inside X, or a numeric free-chat quota, reset interval, or free model. Budget production API use separately: Grok 4.7’s listed short-context text rates are $2 per million input tokens and $6 per million output tokens. At 200,000 prompt tokens, the higher long-context rates apply to every token in the request. (x.ai)