Is Grok 4.5 free?
xAI’s Grok 4.5 announcement advertises “Try for free,” but it does not specify the access route, eligibility, or usage limits. That is not confirmation of a free production API quota: xAI describes API use as usage-based, and the available API information does not establish free production requests. (Grok 4.5 announcement; xAI API)
The API Playground is a separate way to try Grok. xAI says it is included with every account, but the available information does not confirm which models or usage limits apply there. Check the Playground’s model list and terms to see whether Grok 4.5 is available for your account. (xAI API)
Grok 4.5 API pricing
The official model page lists these USD rates per million tokens for requests below its 200,000-token context threshold:
| Token type | Price |
|---|---|
| Input | $2.00 / 1M tokens |
| Cached input | $0.30 / 1M tokens |
| Output | $6.00 / 1M tokens |
The page says requests that exceed the 200K context window have different rates. The pricing information available here does not provide those higher-context rates. The general pricing documentation says that when a prompt reaches a model’s long-context threshold, the long-context rates apply to all tokens in that request. Check the current higher-context schedule before estimating a request whose prompt exceeds 200K tokens. (Grok 4.5 API documentation; API pricing documentation)
The pricing documentation also describes a 1.1× rate for usage sent to a US regional endpoint. The available information does not establish whether that premium applies to Grok 4.5 or which endpoint selection triggers it. Treat the rates above as excluding any regional premium unless you confirm the applicable endpoint terms. The Grok 4.5 page lists us-east-1 and us-west-2 as available regions. (API pricing documentation; Grok 4.5 API documentation)
Illustration below the threshold: 100,000 uncached input tokens and 100,000 output tokens would cost $0.80 at the listed rates: (0.1 × $2) + (0.1 × $6). This rate-based calculation excludes any applicable US regional-endpoint premium; it is not a tested bill. Cached input uses its separately listed rate.
Don’t confuse xAI API rates with Cursor pricing
Cursor separately lists rates for additional Grok 4.5 use beyond its included model pool: $2 input / $6 output per million tokens at standard speed, and $4 / $18 for its Fast variant. These are Cursor-specific rates, not xAI API rates. Cursor says paid plans include a weekly model pool and legacy plans use a monthly pool; its documentation does not say whether the free Cursor plan includes Grok 4.5 or specify the pool size. (Cursor Grok 4.5 documentation)
Before using Grok 4.5 in production
- Confirm the billing channel: a free try or Playground access is not the same as a free production API allowance. For API budgeting, use the xAI rates and your expected input, cached-input, and output token counts.
- Check the documented API model details: xAI lists the model name
grok-4.5, a 500,000-token context window, function calling, structured outputs, and supported reasoning effort levels. These are API details; Cursor’s separate model listing and tools should not be substituted for them. (Grok 4.5 API documentation) - Verify the applicable rate schedule before launch: check higher-context pricing for prompts that exceed 200K tokens, and confirm whether your selected endpoint incurs the documented US regional premium. Current account billing terms and supported regions should also be checked. (API pricing documentation)
Bottom line: xAI advertises a free try for Grok 4.5, but the announcement does not spell out access conditions or limits. A free production API allowance is not confirmed in the available information. For requests below the 200K prompt threshold, the listed rates are $2/M input, $0.30/M cached input, and $6/M output; higher-context requests use different rates, and a US regional-endpoint premium is documented but its applicability to Grok 4.5 remains unverified.