Grok API pricing guide & calculator
Grok API prices range from $1.25 to $2 per million input tokens and $2.50 to $6 per million output tokens, depending on the model. Below: every current model’s price, what common jobs cost per month, and a calculator for your own usage.
Current Grok API prices
| Model | Input | Cached input | Output | Batch in / out |
|---|---|---|---|---|
| Grok 4.7flagship · 500K tokens | $2 | $0.50 | $6 | — |
| Grok 4.3fast · 1M tokens | $1.25 | $0.20 | $2.50 | — |
Per million tokens, standard tier, in US dollars. Checked October 10, 2026 against xAI’s official pricing. “Cached input” is the price for text you reuse across requests, like long instructions.
Estimate your monthly Grok API cost
Pick a job and how many you’ll run a month. Grok’s models are compared with OpenAI and Claude and Gemini; tap a provider to add or remove it. How the estimates work.
| Model | Per ticket | Per month | vs. cheapest |
|---|---|---|---|
| Claude Haiku 5.5Anthropic · fast | $0.00042 | $0.85 | 1× |
| GPT-6 LunaOpenAI · fast | $0.00042 | $0.85 | 1× |
| Gemini 3.5 Flash-LiteGoogle · fast | $0.0016 | $3.25 | 3.8× |
| Gemini 3.8 FlashGoogle · balanced | $0.0032 | $6.38 | 7.5× |
| Grok 4.3xAI · fast | $0.0040 | $8.00 | 9.4× |
| Grok 4.7xAI · flagship | $0.0071 | $14.20 | 17× |
| Claude Sonnet 5.5Anthropic · balanced | $0.0085 | $17.00 | 20× |
| GPT-6.1 SolOpenAI · balanced | $0.0085 | $17.00 | 20× |
| Gemini 3.1 Pro (Preview)Google · flagship | $0.0092 | $18.40 | 22× |
| Claude Opus 5.5Anthropic · flagship | $0.0170 | $34.00 | 40× |
| Claude Fable 5.1Anthropic · flagship | $0.0425 | $85.00 | 100× |
| GPT-6 AstraOpenAI · flagship | $0.0425 | $85.00 | 100× |
Published list prices checked October 10, 2026, before tax. Token counts are estimates and vary by provider, since each counts text a little differently. Subscriptions like Claude Pro, ChatGPT Plus, or Google AI Pro (about $20 a month) are for using the chat apps yourself and can’t be called from your own software. Calculated in your browser; nothing is saved.
Which Grok model should you use?
xAI’s API lineup centers on two current price points. Grok 4.7 is the newest and most capable, at $2 per million input tokens and $6 per million output tokens, with a 500,000 token context window. Grok 4.3 is the cheaper option at $1.25 / $2.50, with a 1 million token context window. xAI also lists reasoning, non-reasoning, and multi-agent variants at Grok 4.3’s price, and a coding model (Grok Build) at $1 / $2.
What stands out is the output price. Output is usually the expensive side of an AI bill, and Grok 4.7’s $6 per million is well below the $10 that OpenAI and Anthropic charge for their middle models. For jobs that produce a lot of text, such as long articles, reports, or detailed answers, that matters more than the input price.
Grok 4.3’s $2.50 output price is lower still, which makes it competitive for writing-heavy, high-volume work, although the budget models from OpenAI and Anthropic are cheaper again for short tasks.
Worked example: a content-heavy business
Say a small marketing agency drafts 100 long articles and 5,000 product descriptions a month. On Grok 4.7 that’s about $13.84 a month. On Claude Sonnet 5.5 or GPT-6.1 Sol, which charge $10 per million output tokens, it’s about $20.24, so Grok 4.7 saves about a third against those middle tiers, almost entirely on the output side, because these jobs write far more than they read. Gemini 3.8 Flash is cheaper still, at about $7.59, while its promotional price lasts, so Grok’s edge is against Claude and OpenAI rather than Google. Whether the cheaper drafts are good enough depends on your brand and how much editing you do; for content a person reviews anyway, a lower price per draft adds up quickly.
What common jobs cost on Grok
Monthly cost at standard prices, with no caching or batch discounts, for typical volumes. The token sizes behind each job are explained in how we calculate.
| Job (per month) | Grok 4.7 | Grok 4.3 |
|---|---|---|
| 1,000 × draft an email | $4 | $2 |
| 2,000 × answer a support ticket | $14.2 | $8 |
| 300 × summarize a 10-page document | $5.46 | $3.15 |
| 100 × write a 1,500-word article | $1.44 | $0.65 |
| 5,000 × write a product description | $12.4 | $6 |
| 3,000 × chatbot conversation (10 turns) | $108 | $60 |
| 200 × ai coding or research agent task | $72 | $42.5 |
Answering 2,000 support tickets a month costs about $14.2 on Grok 4.7 and $8 on Grok 4.3. Writing 100 long articles costs about $1.44 on Grok 4.7, compared with $2.24 on Claude Sonnet 5.5, thanks to the lower output price.
How Grok compares on price
Grok doesn’t have a model in the “balanced” slot the big three use, so the table compares their everyday models. Grok 4.7 ($14.2 for 2,000 support tickets) sits between Gemini 3.8 Flash and the Claude and OpenAI middle tiers. Use the calculator above to see Grok against every model for your own job.
| Balanced model | Input / output per 1M | 2,000 tickets |
|---|---|---|
| Gemini 3.8 FlashGoogle | $0.75 / $3.75 | $6.38 |
| Claude Sonnet 5.5Anthropic | $2 / $10 | $17 |
| GPT-6.1 SolOpenAI | $2 / $10 | $17 |
Discounts: batch, caching, and free use
Cached input. Repeated input costs $0.50 per million on Grok 4.7 instead of $2, and $0.20 instead of $1.25 on Grok 4.3. That’s a smaller discount than the 90% or more some other providers offer, so caching saves less on Grok.
No batch discount listed. xAI’s models page doesn’t list a batch tier, so the calculator’s batch option doesn’t change Grok’s price. Check xAI’s documentation if bulk processing matters to you.
Free tier: Limited free use in the Grok app; no free API tier listed.
What surprises people on the bill
Long prompts double the whole request
When a prompt reaches 200,000 tokens, xAI bills input, cached input, and output at double the normal rate for the entire request. On Grok 4.7, that means $4 / $12 instead of $2 / $6. Our long agent task stays under the threshold, but long-document work can cross it.
Smaller context window on the newest model
Grok 4.7’s context window is 500,000 tokens, half of Grok 4.3’s 1 million and of the 1 million offered by Claude, OpenAI, and DeepSeek’s current models. For most business jobs that’s plenty, but very large document sets may need splitting.
Reasoning variants cost the same but use more tokens
xAI lists reasoning and non-reasoning variants at the same price, but reasoning models produce extra output tokens while they think. The same request can cost more on a reasoning variant.
Prices have moved often
xAI has changed its models and prices several times. Check the official models page before you commit, and recheck periodically.
Grok API vs. SuperGrok Lite
xAI’s consumer plans and the API are separate. SuperGrok (about $30 a month) is for using Grok in its app, with cheaper and more expensive tiers available. The API is for building Grok into your own software, billed per token.
For a sense of scale, $20 of API usage buys about 5,000 email drafts on Grok 4.7 or 10,000 on Grok 4.3, using our token estimates.
| xAI subscription | Price |
|---|---|
| SuperGrok Lite | $10/mo |
| SuperGrok | $30/mo |
| SuperGrok Heavy | $300/mo |
Who the Grok API is a good fit for
Grok’s API is worth testing for output-heavy jobs, where its low output price makes the biggest difference: long articles, reports, detailed answers, and content generation. Grok 4.3’s 1 million token context window at a low price also suits long-document reading.
The trade-offs are a smaller caching discount, no listed batch tier, and frequent model changes. As with every provider, run your real task on a few models and pick the cheapest one whose results you’d be happy to send.
Frequently asked questions
How much does the Grok API cost?
Grok 4.7 costs $2 per million input tokens and $6 per million output tokens, and Grok 4.3 costs $1.25 / $2.50. Requests whose prompt reaches 200,000 tokens are billed at double those rates for the whole request.
Is the Grok API cheaper than OpenAI?
Compared with OpenAI’s middle model, GPT-6.1 Sol at $2 / $10, Grok 4.7’s $2 / $6 is cheaper on output, which usually makes it cheaper overall. OpenAI’s budget model, GPT-6 Luna, is cheaper than either Grok model for short tasks.
Does SuperGrok include API access?
No. SuperGrok is a subscription for the Grok app. The xAI API is billed separately, per token.
What is Grok’s context window?
Grok 4.7 has a 500,000 token context window, and Grok 4.3 has 1 million tokens, according to xAI’s models page.
Does xAI offer a batch discount?
xAI’s models page doesn’t list a batch tier. Cached input is discounted: $0.50 per million on Grok 4.7 and $0.20 on Grok 4.3.