xAI · AI model API pricing

How much does Grok 4.5 cost?

xAI's flagship — by far the cheapest output tokens of any frontier-class model ($6 vs $25–50 elsewhere).

Rate card

What that means per month

Standard on-demand rates from xAI's official model docs, last checked 2026-07-28. Rates shown are for prompts under 200k tokens; above that, prices double. xAI offers up to $175/month in free credits via its data-sharing program. Monthly figures exclude caching and batch discounts. Confirm live pricing before committing.

Where Grok 4.5 sits on price

At the reference month of 10M input and 3M output tokens, Grok 4.5 costs $38, which makes it the 21st cheapest of the 30 models tracked here. That is 22× the price of Qwen-Flash, the cheapest model at this mix ($1.70). Among the 10 frontier-class models it ranks 4th, behind GLM-5.2 at $27.2.

Cache economics

Cached input on Grok 4.5 bills at $0.30 per 1M — 15% of its own input rate, against a 10% median across the models here that publish a cached-read price. 20 models discount cached reads more steeply than this. On the reference month, serving 80% of that input from cache takes the input side of the bill from $20 to $6.40, leaving output untouched at $18 — which is why a high reuse rate changes the ranking above rather than just shaving the total.

Cache reads only. Writes are billed separately by some providers and are excluded here — see the rate note above for xAI's terms. Try it at your own hit rate on the AI cost calculator.

Cheaper alternatives

At a reference workload of 10M input / 3M output tokens a month, these cost less than Grok 4.5 ($38/mo):

• Grok 4.6 $38/mo
• Mistral Medium 3.5 $37.5/mo
• Qwen3-Max $30/mo

Grok 4.5 head-to-head