xAI Grok API Pricing 2026: Token Costs, SuperGrok Plans, and Full Comparison
Exact xAI Grok API pricing for 2026 -- Grok 4.3, Grok 3, Grok 3 Mini, and Grok 4.1 Fast token rates, free developer credits, SuperGrok subscription plans, and how Grok API compares to OpenAI and Anthropic.
Victor OgonyoIf you are searching for xAI Grok API pricing in 2026, you are in the right place. This guide covers every token rate for every Grok model, the $175/month free developer credit program, all four SuperGrok consumer subscription tiers, and a direct comparison of Grok API pricing against OpenAI and Anthropic's APIs.
Whether you need the Grok API for a production application or you are evaluating Grok AI pricing for a personal subscription, this breakdown covers everything with exact numbers.
Grok Pricing at a Glance
| Product | Price | Best For |
|---|---|---|
| Free (on X) | $0 | Casual, limited use |
| SuperGrok Lite | $10/month | Image/video gen, light AI use |
| SuperGrok | $30/month ($300/year) | Grok 4.3, DeepSearch, Think mode |
| X Premium+ | $40/month | Grok + ad-free X bundle |
| SuperGrok Heavy | $300/month | Max context, max rate limits |
| xAI API | Pay per token | Developers building applications |
xAI Grok API Pricing 2026: Token Rates by Model
The xAI API is the developer access point for all Grok models. Pricing is per million tokens (input and output billed separately).
Grok API Rates (Per Million Tokens)
| Model | Input | Output | Context Window |
|---|---|---|---|
| Grok 4.3 | $1.25 | $2.50 | 1M tokens |
| Grok 3 | $1.25 | $2.50 | 131K tokens |
| Grok 3 Mini | $0.30 | $0.50 | 131K tokens |
| Grok 4.1 Fast | $0.20 | $0.50 | 131K tokens |
| Cache Read | $0.05 | — | — |
What Each Grok API Model Is For
Grok 4.3 is xAI's flagship model via API. At $1.25/$2.50 per million tokens with a 1M token context window, it is the most capable and context-rich option. A 100,000-token document costs $0.125 to process — competitive with Claude Sonnet 4.6 and significantly cheaper than GPT-4o.
Grok 3 offers the same $1.25/$2.50 rate as Grok 4.3 but with a 131K context window instead of 1M. Use it when you do not need the extended context but want frontier model quality.
Grok 3 Mini at $0.30/$0.50 is the mid-tier option — capable enough for summarisation, classification, and structured extraction at roughly a quarter of the flagship model's cost.
Grok 4.1 Fast at $0.20/$0.50 is xAI's cheapest capable model. At $0.20 per million input tokens, it competes directly with GPT-4o Mini as a high-volume, lower-complexity workhorse.
Cache Read at $0.05 per million tokens lets you reuse cached context (system prompts, large documents) for a fraction of the input rate. This is the same prompt caching concept Anthropic offers and dramatically reduces costs for applications with repeated context.
Free xAI API Developer Credits
All xAI API developers receive $175 per month in free credits. For small applications, prototypes, or testing, this means genuinely free API access within the credit limit. At $1.25/M input tokens on Grok 4.3, $175 covers approximately 140 million input tokens per month before any charges apply.
Grok API vs OpenAI vs Anthropic: Price Comparison
Here is how Grok API pricing compares to the equivalent models from OpenAI and Anthropic at identical capability tiers.
Low-Cost Models
| Model | Input (per 1M) | Output (per 1M) |
|---|---|---|
| Grok 4.1 Fast | $0.20 | $0.50 |
| GPT-4o Mini | $0.15 | $0.60 |
| Grok 3 Mini | $0.30 | $0.50 |
| Claude Haiku 4.5 | $1.00 | $5.00 |
Grok 4.1 Fast undercuts Claude Haiku 4.5 significantly. It costs more than GPT-4o Mini on input but less on output — net cost depends on your output-to-input ratio.
Frontier Models
| Model | Input (per 1M) | Output (per 1M) |
|---|---|---|
| Grok 4.3 / Grok 3 | $1.25 | $2.50 |
| GPT-4o | $2.50 | $10.00 |
| Claude Sonnet 4.6 | $3.00 | $15.00 |
| Claude Opus 4.7 | $5.00 | $25.00 |
Grok 4.3 is the most affordable frontier model in this comparison. At $1.25/$2.50, it is half the price of GPT-4o on input and one-sixth the price of Claude Opus 4.7 on output. For output-heavy applications (long-form generation, detailed analysis), the output cost gap is especially significant.
The trade-off: Grok's benchmark performance on complex reasoning tasks is competitive but varies by task type. GPT-4o and Claude Opus 4.7 outperform Grok 4.3 on certain code generation and multi-step reasoning benchmarks. For applications where reasoning quality is the bottleneck, test before committing.
SuperGrok Consumer Plans: Full Breakdown
For non-developers who want Grok as an AI assistant, xAI offers four subscription tiers.
Free Tier
Anyone with an X account gets basic Grok access for free.
Free tier includes:
- Approximately 10 prompts per 2-hour rolling window
- Limited web search via real-time X and web data integration
- Access to a lighter Grok model
- Restricted image generation
Grok's free tier is more limited than Claude's free tier and ChatGPT's free access. The unique advantage is real-time X data — even free users can query Grok about trending topics, recent X posts, or breaking news. No other AI assistant offers this natively.
SuperGrok Lite — $10/Month
SuperGrok Lite launched in March 2026 as an entry-level paid tier.
SuperGrok Lite includes:
- Basic AI image generation (Aurora model)
- Basic AI video generation
- Access to one AI agent
- Higher message limits than free (exact cap not publicly specified)
- Standard Grok model access (not Grok 4.3)
Best for: Casual users who primarily want AI-generated images or short video clips and need more headroom than the free tier provides.
SuperGrok — $30/Month ($300/Year)
SuperGrok is xAI's primary consumer product and the direct competitor to Claude Pro ($20/month) and ChatGPT Plus ($20/month).
SuperGrok includes:
- Grok 4.3 — xAI's flagship model
- 128K token context window
- DeepSearch — real-time web and X search integrated into responses
- Big Brain mode — extended compute reasoning for complex problems
- Think mode — step-by-step reasoning shown before answers
- Voice mode
- Unlimited AI image generation (Aurora model)
- Approximately 30 messages per 2-hour window on Grok 4.3
The X data advantage: SuperGrok's defining capability is DeepSearch with live X integration. For journalists, PR professionals, market researchers, political analysts, and social media managers, this is a genuine capability gap over competitors. No other AI assistant provides this depth of real-time social platform data.
The price premium: At $30/month, SuperGrok costs 50% more than Claude Pro and ChatGPT Plus. The extra $10/month primarily pays for the real-time X data integration. If that is not valuable to your work, Claude Pro or ChatGPT Plus delivers comparable general reasoning at a lower price.
Message limits: The 30-message-per-2-hour cap on Grok 4.3 is more restrictive than ChatGPT Plus (approximately 150 messages per 3 hours on GPT-4o). Heavy users will hit it.
X Premium+ — $40/Month
X Premium+ bundles SuperGrok capability with an ad-free X experience.
X Premium+ includes:
- Full Grok 4 access (same capability as SuperGrok, slightly older model than Grok 4.3)
- Ad-free X browsing
- X Premium+ account badge and features
- Creator revenue sharing eligibility
- Priority customer support on X
Best for: Heavy X users who want both an AI assistant and a better X experience. If you spend significant time on X and would pay for the ad-free experience separately, Premium+ bundles the value efficiently at $40/month.
Skip it if: You primarily want the AI capability and use X infrequently. SuperGrok at $30/month gives you the same AI features for $10 less.
SuperGrok Heavy — $300/Month
SuperGrok Heavy is xAI's premium tier for professionals requiring maximum capability.
SuperGrok Heavy includes:
- Full Grok 4.3 with parallel compute (test-time compute scaling)
- 256K token context window (double the standard SuperGrok)
- Maximum rate limits — significantly more messages per hour than standard SuperGrok
- Priority processing
- Full access to Grok's most compute-intensive reasoning modes
Best for: Researchers, developers, and enterprise professionals running Grok on book-length documents, large codebases, or extended multi-step analysis sessions where both the context size and message throughput matter. The 256K context window is the largest offered in any Grok consumer tier.
Cost reality check: At $300/month — 10x the cost of standard SuperGrok — this is for organisations or power users who consume Grok at very high volume and need both the extended context and the throughput headroom.
Grok's Key Features Explained
DeepSearch
DeepSearch is Grok's integrated real-time web and X search tool. Unlike ChatGPT's web browsing or Perplexity's search results, DeepSearch sources live X posts as primary data alongside broader web crawl.
Unique use cases this enables:
- Tracking breaking news as it develops on X in real time
- Understanding current public sentiment on a brand, political topic, or event
- Researching what practitioners are saying in specific technical communities right now
- Monitoring competitor announcements and community reaction
No competitor replicates this. It is Grok's clearest product differentiation.
Think Mode and Big Brain Mode
Think mode displays Grok's step-by-step reasoning before the final answer — equivalent to Claude's extended thinking or GPT-4o's reasoning mode. Useful for complex problems where transparency into reasoning matters.
Big Brain mode allocates additional compute for harder problems: multi-step mathematics, complex code generation, nuanced analysis with multiple constraints. Available on SuperGrok and above, at the cost of slower responses.
Aurora Image Generation
Grok's image generation uses xAI's Aurora model. SuperGrok includes unlimited image generation with no extra cost. Quality is competitive with DALL-E 3, with particular strength in photorealistic and stylised outputs.
Grok vs Claude Pro vs ChatGPT Plus
| Feature | Grok SuperGrok | Claude Pro | ChatGPT Plus |
|---|---|---|---|
| Monthly price | $30 | $20 | $20 |
| Flagship model | Grok 4.3 | Claude Opus 4.7 | GPT-4o |
| Context window | 128K | 100K+ | 128K |
| Real-time web search | Yes (DeepSearch) | No | Yes |
| Live X/social data | Yes (unique) | No | No |
| Image generation | Yes (unlimited) | No | Yes (DALL-E, limited) |
| Voice mode | Yes | Yes | Yes |
| Extended reasoning | Yes (Big Brain) | Yes (extended thinking) | Yes |
| Code execution | Limited | Yes (Claude Code) | Yes (Code Interpreter) |
| Message limits | ~30/2hrs on Grok 4.3 | Higher (rolling) | ~150/3hrs |
| API available | Yes | Yes | Yes |
Grok's advantages: Real-time X platform data, unlimited image generation at no extra cost, frontier API pricing (Grok 4.3 at $1.25/$2.50 vs GPT-4o at $2.50/$10.00).
Claude's advantages: Claude Code for software development, extended thinking quality, writing and reasoning benchmarks on complex tasks, prompt caching savings.
ChatGPT's advantages: Price per message at the Plus tier ($20 vs $30), Code Interpreter for data analysis, broader third-party integrations and plugin ecosystem.
Who Should Use Grok and the xAI API
Choose SuperGrok ($30/month) if:
- You work in journalism, PR, politics, or social media and need real-time public sentiment data
- You want unlimited AI image generation without paying extra
- You are already an X Premium subscriber
- You need a 128K context window for processing longer documents
Choose the xAI API if:
- You are building an application and want frontier AI at lower output cost than GPT-4o ($10/M) or Claude Sonnet 4.6 ($15/M)
- You want $175/month in free credits for prototyping
- You need a 1M token context window (Grok 4.3) for long-document processing at $1.25/M input
- You need a high-volume low-cost model (Grok 4.1 Fast at $0.20/$0.50)
Skip Grok if:
- You primarily need software development assistance — Claude Code is better
- You want more messages per hour — ChatGPT Plus is more generous
- You do not use X/Twitter and do not need real-time social data
- Budget is tight — Claude Pro and ChatGPT Plus offer comparable reasoning at $10/month less
Frequently Asked Questions
What is xAI Grok API pricing in 2026? Grok 4.3 costs $1.25 per million input tokens and $2.50 per million output tokens via the xAI API. Grok 4.1 Fast costs $0.20/$0.50, Grok 3 Mini costs $0.30/$0.50. All developers receive $175/month in free credits.
Is the Grok API free? The xAI API offers $175/month in free credits, which covers substantial usage on lower-cost models. Beyond that, you pay per token at the rates listed above.
How much does SuperGrok cost? SuperGrok costs $30/month or $300/year. SuperGrok Lite is $10/month. SuperGrok Heavy is $300/month. X Premium+ (which includes Grok) is $40/month.
Is Grok API cheaper than OpenAI? For frontier models, yes. Grok 4.3 at $1.25/$2.50 per million tokens is significantly cheaper than GPT-4o at $2.50/$10.00. For low-cost models, GPT-4o Mini at $0.15/$0.60 is cheaper than Grok 4.1 Fast at $0.20/$0.50 on input but more expensive on output.
What is the context window for Grok API models? Grok 4.3 has a 1M token context window via API. Grok 3, Grok 3 Mini, and Grok 4.1 Fast have a 131K token context window. SuperGrok Heavy (consumer subscription) includes a 256K context window.
How does Grok API compare to Claude API? Grok 4.3 ($1.25/$2.50) is significantly cheaper than Claude Sonnet 4.6 ($3.00/$15.00) and Claude Opus 4.7 ($5.00/$25.00). Claude offers prompt caching (up to 90% savings) and batch processing (50% discount), which can bring effective costs lower for certain workloads. Claude also has stronger benchmark performance on complex reasoning and coding tasks.
The Bottom Line
xAI Grok API pricing in 2026 is genuinely competitive, especially at the frontier model tier. Grok 4.3 at $1.25/$2.50 per million tokens undercuts GPT-4o and Claude Sonnet 4.6 by a significant margin. The $175/month free developer credits lower the barrier to entry further. For applications where Grok's reasoning quality meets the bar and real-time X data integration adds value, the xAI API is one of the better-priced options in the market.
The consumer SuperGrok plans are well-structured, though the $30/month price point (50% above Claude Pro and ChatGPT Plus) is only justified if DeepSearch and the X data integration are genuinely valuable to your specific workflow.
Building an AI startup? List it on Startup Launch Page and reach investors and early adopters actively looking for what you're building.
Building something great?
List your startup on Startup Launch Page -- reach real investors, founders, and early adopters.
Launch your startup →