Grok AI Capabilities 2026: Everything Grok Can Do, Features & Access Guide

Last Updated: November 1, 2026

Grok AI can chat, reason, write code, search the web and X in real time, generate images and videos, hold voice conversations, and run AI agents that finish multi-step work through Grok Bot. As of November 2026, Grok's flagship model is Grok 4.7, with a 500,000-token context window and API pricing of $2 per million input tokens and $6 per million output tokens, according to the official Grok 4.7 documentation.

Grok is built by SpaceXSI (SpaceX Super Intelligence), the AI company formerly known as xAI. SpaceX announced it had acquired xAI on February 2, 2026, according to the company's news page, and the footer on x.ai now reads SpaceX Super Intelligence LLC.

Since October 12, 2026, new customers buy Grok through The X Subscription, one plan that bundles Grok, the Cursor coding agent and X perks, priced from free to $200 per month (official announcement). The table below summarizes every major Grok capability, what it does, and where you can use it today.

This guide breaks down each Grok capability, which Grok model fits which job, what each plan costs, and where Grok still trails ChatGPT and Claude for business use.

🎯 Before you read on, we put together a free 2026 AI Tools Cheat Sheet covering the tools business leaders are actually using right now. Get it instantly when you subscribe to AI Business Weekly.

Get the Free Cheat Sheet

Before the capability breakdown, three misconceptions about Grok are worth clearing up because they change how you evaluate everything else.

Misconception 1: Grok is just a political AI built for X users. Grok launched around Elon Musk's stated goal of a less politically restricted AI, and some early versions created real controversy. The 2026 product is aimed squarely at work: SpaceXSI describes Grok 4.7 as its frontier model "built for coding, agentic tasks, and knowledge work," and in July 2026 it announced Grok for Excel, Grok for Outlook and Grok in Google Workspace on its news page.

Misconception 2: Grok is only useful for X-related content. Real-time X search is Grok's most distinctive feature, not its only one. Grok also handles long documents (up to 1 million tokens on Grok 4.3 and Grok 4.20), image and video generation, voice, coding through Grok Build, and delegated multi-step work through Grok Bot.

Misconception 3: Grok's controversies make it unsuitable for business. The content controversies of 2024 and 2025 were real, and they centered on consumer features such as image generation and companions. Business teams usually evaluate Grok through the API or through Grok Bot, whose enterprise version is still a waitlist. Knowing which Grok product you are evaluating matters before you dismiss the platform.

Grok answers questions about current events by calling live search tools, including web search and X search, while it reasons. The Grok 4.7 documentation lists web search, X search and code execution as built-in tools on the API.

Most AI assistants now search the web, so live information alone no longer sets Grok apart. Grok's edge is its X search tool, which searches posts on X directly. Breaking news, executive commentary and customer complaints often surface on X hours before they appear in articles that web search engines can crawl.

Grok's X search is most useful for questions such as: what are people saying about a product launch in the last few hours, which accounts are driving a story, and how sentiment shifted after an announcement. A web-only search tool sees those conversations later, if at all, once someone writes them up.

The trade-off with Grok's real-time answers is quality control. Posts on X are fast but unverified, so Grok's real-time answers are only as reliable as the posts it finds. For anything you will publish or act on, treat Grok's X findings as leads to confirm, not as settled facts.

For journalists, market researchers, competitive intelligence teams and social media managers, Grok's X search remains the clearest reason to add Grok to the toolkit.

Grok's main features in 2026 are frontier reasoning models, real-time web and X search, AI agents through Grok Bot, coding through Grok Build, image and video generation through Grok Imagine, and voice. Here is what each Grok feature does.

Grok 4.7 is the current flagship Grok model. SpaceXSI's developer documentation says Grok 4.7 takes text and image input, returns text, and offers four reasoning levels: low, medium, high (the default) and xhigh. Its knowledge cutoff is May 2026. Grok 4.7 launched in September 2026 and scored 46 on the independent Artificial Analysis Intelligence Index, two points above Grok 4.6, ForkLog reported.

Grok's web search and X search tools let it pull current information into an answer instead of relying only on training data. On the API, Grok 4.7 can use these search tools alongside function calling and code execution, per the Grok 4.7 docs.

Grok Bot, introduced on August 11, 2026, gives you AI teammates that run on their own cloud computer, sign into the tools you already use, and handle multi-step jobs across apps, inboxes and websites. Grok Bot contacts you only when something needs approval, remembers past conversations, and can learn a workflow you demonstrate, according to SpaceXSI's launch post. Grok Bot is in beta on desktop and iOS, and its usage is metered separately from Grok and Cursor plan usage (SpaceXSI).

Grok Build is SpaceXSI's coding agent, available on web and mobile, and Grok 4.7 is its default model. Grok 4.7 is also offered on every Cursor plan, and a faster variant, Grok 4.7 Fast, runs only inside Cursor and Grok Build at twice the standard token rate, according to the Grok 4.7 documentation.

Grok supports spoken conversation, and voice is listed as a core Grok feature on The X Subscription announcement. For developers, SpaceXSI's Voice API charges $0.08 per minute for speech-to-speech, $15 per million characters for text-to-speech, and $0.10 per hour for batch speech-to-text, according to its models and pricing page.

Imagine Image 2.0, released in August 2026, is Grok's current image model. Imagine Image 2.0 adds Magic Wand editing of a selected area, background removal, multi-reference editing that combines up to five images, and Smart Resize to any aspect ratio. In Arena rankings as of August 7, 2026, its faster variant placed second in both text-to-image and image editing, behind OpenAI's GPT-Image-2, The Decoder reported. On the API, Grok image generation starts at $0.02 per image (SpaceXSI pricing).

Imagine Video 1.5 is Grok's current video model. Imagine Video 1.5 supports text-to-video, image-to-video, native 1080p output, up to seven image references to lock in a face, product or location, and voice references that keep a character's voice consistent across scenes, according to SpaceXSI's July 31, 2026 announcement. API video pricing starts at $0.02 per second. Our Grok video generation guide covers prompts, limits and use cases in more depth.

Grok's context window depends on the model. Grok 4.7 handles 500,000 tokens, while Grok 4.3 and Grok 4.20 each handle 1,000,000 tokens, according to SpaceXSI's Grok 4.3 and Grok 4.20 model pages. At the common rule of thumb of about 0.75 words per token, 1 million tokens is roughly 750,000 words, enough for a large contract set or a mid-sized codebase in one request. Our Grok context window guide explains the limits in more detail.

Grok now reaches into everyday office software. In July 2026, SpaceXSI announced Grok for Excel, Grok for Outlook, Grok in Google Workspace and Automations in Grok, according to its news page.

Most people should use Grok 4.7, the flagship Grok model, for hard reasoning, coding and agent work. Developers who need the largest context window or a lower price should look at Grok 4.3 or Grok 4.20, which both offer 1 million tokens for less, according to SpaceXSI's models and pricing page and individual model pages.

Grok API costs scale with tokens, so model choice matters. Take a team that sends 10 million input tokens and receives 2 million output tokens in a month. On Grok 4.7, input costs $20 and output costs $12. On Grok 4.3, input costs $12.50 and output costs $5. On our math, the same monthly workload costs $32 on Grok 4.7 and $17.50 on Grok 4.3, so the flagship carries roughly an 83% premium for its extra capability.

Those Grok cost figures use list prices for standard-length requests. Grok 4.7 cached input costs $0.50 per million tokens, and requests above 200,000 tokens are billed at double the standard rate, ForkLog reported, so long-context jobs can cost more than the simple math suggests. For every plan and API price in one place, see our Grok AI pricing breakdown.

SpaceXSI retired Grok 4 (grok-4-0709), Grok 4 Fast, Grok 4.1 Fast, grok-code-fast-1 and Grok 3 from its API on May 15, 2026. Most calls to those retired Grok models now redirect to Grok 4.3, and grok-code-fast-1 moved to grok-build-0.1, according to the official migration guide.

Teams that built workflows on Grok 4 or Grok 4.1 Fast earlier in 2026 should confirm those calls now point to a current Grok model on purpose rather than through a redirect. Output quality and pricing can differ from the model you originally tested.

💡 Finding this helpful? Get bite-sized AI news and practical business insights like this delivered free every morning at 7 AM EST.

Subscribe Free

Grok leads on live X data, long context and image and video tools, and it trails ChatGPT and Claude on brand trust and enterprise maturity. The table gives our verdict by area.

Where Grok genuinely leads: Grok's X search is the feature rivals find hardest to copy, because the freshest signal on many business stories lives in X posts. Grok also bundles a lot into one bill: The X Subscription combines Grok, Cursor and X perks, and SpaceXSI says its $30 Premium tier includes benefits previously sold for $90 per month combined (SpaceXSI).

Where Grok falls short: Grok's enterprise offering is younger than OpenAI's or Anthropic's, with Grok Bot for Enterprise still on a waitlist. Reputation concerns from Grok's earlier controversies still slow adoption in regulated industries. For a head-to-head on features, pricing and output, read our Grok vs ChatGPT comparison.

New customers get Grok through The X Subscription, which launched on October 12, 2026 with five tiers. Existing SuperGrok subscribers can keep their current plan, price and features, or switch and receive a prorated refund for the old plan, according to SpaceXSI's announcement.

Which Grok plan fits: Casual users can start on Free or Plus. Premium at $30 is the first tier that includes SuperGrok's benefits, so Premium is the realistic starting point for daily professional use. Super and Ultra mainly raise usage to 15x and 50x the Plus allowance, so they only make sense for heavy users who regularly hit limits.

Where you can use Grok: Grok runs on grok.com, the iOS and Android apps, and inside X. On paid X Subscription plans, Grok and Cursor draw from one shared pool of usage.

Legacy SuperGrok plans: SuperGrok, SuperGrok Plus and SuperGrok Heavy subscribers can stay on their plans. All three legacy SuperGrok plans include Grok Bot, with Bot usage metered separately, per SpaceXSI.

SpaceXSI API: Developers access Grok through the API documented at docs.x.ai. Grok 4.7 costs $2 per million input tokens and $6 per million output tokens, and a US regional endpoint is priced at a 10% premium. Grok 4.7 is also available through the OpenRouter, Vercel and Cloudflare model gateways, per the Grok 4.7 docs.

Grok Bot for Enterprise: Larger companies can join a waitlist for the enterprise version of Grok Bot, according to SpaceXSI's Grok Bot launch post.

Grok delivers the most business value where live information, long documents or delegated multi-step work matter. These are the strongest Grok use cases in 2026.

Market and Competitive Intelligence: Grok's X search makes it a strong tool for monitoring market sentiment, tracking competitor announcements as they happen, and spotting emerging narratives before they reach news coverage. Teams that do this work manually can replace much of the scrolling with targeted Grok prompts.

Research and Document Analysis: Grok 4.3 and Grok 4.20 accept up to 1 million tokens per request, which suits legal, consulting and policy teams working with large document sets.

Social Media and Communications: Real-time sentiment monitoring, trend identification, and content drafting for X specifically are natural Grok use cases. Comms teams that manage significant X presence find Grok more useful than ChatGPT for this function.

Content Creation with Current Context: Content teams that reference current events or trends can use Grok's live web and X search to gather that context inside the same chat where they draft.

Delegated Multi-Step Work: Grok Bot can sign into business tools and complete multi-step jobs across apps, inboxes and websites, asking for approval when needed. Grok Bot is still in beta, so start with low-risk workflows before handing it anything customer-facing.

Software Development: Grok Build and Grok 4.7 target coding work directly, and Grok 4.7 is available on every Cursor plan. For a complete picture of AI coding tools, our AI coding tools guide covers the full comparison.

For an overview of how to choose the right AI tool for each business function across your organization, our AI for business guide covers the full decision framework.

What is Grok AI? Complete Guide 2026: The full background on Grok, including its founding story, model history, and how Grok evolved from its controversial early versions to the 2026 platform.

Grok AI Statistics 2026: Users, Growth & Market Share: The numbers behind Grok, including user counts, market share and growth trajectory data.

What is xAI? Complete Guide 2026: The company behind Grok, covering its founding, the Colossus supercomputer, the SpaceX acquisition and its strategic position in the AI industry.

Best AI Chatbots for Business 2026: How Grok compares to ChatGPT, Claude, Gemini and Perplexity across specific business use cases, with recommendations.

AI Chatbots Comparison Guide 2026: Full feature and pricing comparison across all major AI chatbot platforms, including Grok, ChatGPT, Claude and Gemini.

What can Grok AI do in 2026? Grok can chat and reason with the Grok 4.7 model, search the web and X in real time, generate images and videos with Grok Imagine, hold voice conversations, write code through Grok Build, and complete multi-step tasks as an AI agent through Grok Bot. On the API, Grok context windows reach 1 million tokens on Grok 4.3 and Grok 4.20.

What is the latest Grok model? Grok 4.7 is SpaceXSI's flagship model as of November 2026, with a 500,000-token context window and API pricing of $2 per million input tokens and $6 per million output tokens. Grok 4.3 and Grok 4.20 remain available with 1 million token context windows, while Grok 4 and Grok 4.1 Fast were retired from the API on May 15, 2026.

How much does Grok cost? New customers buy Grok through The X Subscription: Free, Plus at $8 per month, Premium at $30, Super at $100 and Ultra at $200. Existing SuperGrok subscribers can keep their legacy plans or switch and receive a prorated refund.

Can Grok generate images and videos? Yes. Grok Imagine creates images with Imagine Image 2.0 and videos with Imagine Video 1.5, which supports text-to-video, image-to-video and native 1080p output. On the API, Grok image generation starts at $0.02 per image and video starts at $0.02 per second.

Who owns Grok now? Grok is made by SpaceXSI (SpaceX Super Intelligence), the company formerly known as xAI. SpaceX announced its acquisition of xAI on February 2, 2026, and the x.ai website footer now reads SpaceX Super Intelligence LLC.

The Grok facts in this guide were checked on November 1, 2026 against SpaceXSI's official pages: the models and pricing page, the Grok 4.7 documentation, the May 15, 2026 model retirement guide, and The X Subscription announcement. Benchmark and ranking figures come from ForkLog and The Decoder, as linked in each section.

Grok in 2026 is a broader and more business-ready product than its reputation suggests. Real-time X search, 1 million token context on Grok 4.3 and Grok 4.20, Grok Bot agents, and strong image and video tools give Grok genuine advantages in specific workflows.

Grok is not the right tool for every workflow. Grok's enterprise maturity still trails OpenAI and Anthropic, and its reputation in regulated industries remains a friction point that procurement and legal teams will raise.

The most useful framing is not "Grok vs ChatGPT" but "Grok for live social intelligence, long documents and agent experiments, ChatGPT or Claude for everything else." For a side-by-side comparison of Grok, ChatGPT, Claude and Gemini across business workflows, our best AI chatbots for business guide covers the full decision framework.

📨 Don't miss tomorrow's edition. Subscribe free to AI Business Weekly and get our 2026 AI Tools Cheat Sheet instantly, plus bite-sized AI news every morning with zero hype.

Subscribe Free

Written by the AI Business Weekly Editorial Team. Reviewed, edited and published by Sameer Khan. This article was AI-assisted, then reviewed by Sameer Khan before publishing.