Grok API pricing in 2026: xAI token, tool and media rates

How much does each Grok model cost per million tokens?

xAI lists one rate for short prompts and a higher rate once a prompt reaches 200,000 tokens. The long-context rate then applies to every token in that request. Figures come from xAI’s API pricing page. Prices checked November 10, 2026.

Grok (xAI) and Groq are different companies. Groq runs open models on its own chips and has its own pricing.

What changes the price of a Grok request?

Several options raise or lower the token price. Server-side tools add a fee per call or per item on top of tokens, and the model decides how many tools to call, so cost grows with query complexity.

• Batch API: 20% off all token types on Grok 4.7, 4.6, 4.5, 4.3 and the 4.20 models. Most batches finish within 24 hours and do not count toward rate limits. Other models, including grok-build-0.1, get no batch discount.
• Priority processing: 2 times standard rates, billed only when the response confirms the priority tier.
• US regional endpoint: 1.1 times standard rates. It serves Grok 4.7 and 4.6 only.
• Policy violations: a request blocked before generation in the Responses API costs $0.05.
• Grok 4.7 Fast exists only inside Cursor and Grok Build, not on the public API.

What do Grok images, video and voice cost?

xAI prices media per output. Image editing inputs also cost $0.002 to $0.01 per image. Batch requests for images and video bill at standard rates.

What does a real Grok workload cost?

Take 10,000 calls, each with a 2,000-token prompt and a 500-token answer: 20 million input tokens and 5 million output tokens. The cached column assumes 1,500 tokens of each prompt are a cache hit.

Reasoning tokens bill as tokens too, so a model that thinks before answering costs more than the table shows. If every call also runs one web search, add $50 for the 10,000 searches.

Does the Grok API have a free tier or rate limits?

xAI’s pricing page lists no free tier. Rate limits rise with cumulative API spend since January 1, 2026, and a team never drops a tier. Buyers of X API credits can earn up to 20% back in xAI API credits, as described in the X API pricing guide.

• X API pricing

How does Grok work through Locus Pro?

Locus Pro offers Grok chat, web search, X search, code execution, speech and images as separate endpoints, paid from one prepaid credit balance with no xAI account. Catalog checked November 10, 2026.

grok/chat accepts grok-4.7, grok-4.6, grok-4.5, grok-4.3 and the two grok-4.20-0309 models. The multi-agent model is excluded because it does not accept an output cap, and max_tokens is required. Each call holds an amount sized from the prompt and output budget, then settles to reported usage.

• Add Locus to Grok
• Browse model endpoints

Which Grok model and setup should you choose?

For steady volume, Grok can cost less per call with your own xAI account and API key, which also gets the Batch discount and rate limits that rise with spend.

Locus Pro fits an agent that uses Grok for one job, such as X search, next to other models and data APIs. You get one balance, a max_credits ceiling on each call, a monthly spend limit and no separate xAI billing.

• Best quality and agentic tool calling: grok-4.7.
• Long documents up to 1 million tokens at lower cost: grok-4.3 or the grok-4.20 models.
• Cheapest text rates: grok-build-0.1 at $1 and $2.
• Searching X posts: compare the X Search tool at $5 per 1,000 posts plus tokens with the X API at $0.005 per post read.