Grok vs Claude: the models, apps, coding tools and plans, from the official docs
Updated on
Grok vs Claude is really two families of products: xAI's Grok models, apps and tools, and Anthropic's Claude models, apps and tools. This page puts them side by side using only what the two companies publish: the current models, context windows, live search, image and video, coding agents, and how plans and usage limits work.
Status: Not tested yet. We have not run Grok and Claude on the same job, so this page has no winner and no scores of our own. Sources were checked on 2026-10-25, and both companies change these pages often.
The short version
Grok is xAI's assistant on grok.com and in the iOS and Android apps. It chats, makes images and videos with Grok Imagine, talks by voice, reads uploaded files, and connects to your email, files and calendar. xAI's API offers the same model family to developers, and Grok Build is its coding agent for the terminal.
Claude is Anthropic's assistant on the web, desktop, iOS and Android. Every plan, including Free, lists web search, file creation, running code, memory, connectors and skills. Claude Code, Anthropic's coding agent, comes with the paid plans, and the Claude API offers the models to developers.
The clearest differences in the official pages: Grok has live X data and its own image and video generator; Claude's current models have a larger context window on the API and output text only.
Side by side
| Grok (xAI) | Claude (Anthropic) | |
|---|---|---|
| Model the company points you to | Grok 4.7, which xAI calls its most capable model and recommends for both code and chat. | Claude Opus 5.5 for most work, and Claude Fable 5.1 for demanding reasoning and long agentic work. Sonnet 5.5 and Haiku 5.5 are the faster options. |
| Context window (API) | 500k tokens for Grok 4.7. Some older Grok models are listed at 1M. | 1M tokens on all four current models, with up to 128K tokens of output. |
| Knowledge cutoff | May 2026 for Grok 4.7. | June 2026 for the current models. |
| Live information | The app lists real-time web search and live X integration. On the API, Grok has no current data unless you turn on the Web Search or X Search tools. | Web search on every plan, and a web search tool on the API. No X-specific search is listed. |
| Images and video | Grok Imagine makes images and videos; the product page lists text-to-video up to 15 seconds at 720p. Separate Imagine models are on the API. | The current models take text and images in and give text out. Claude Motion, a beta on Team and Enterprise plans, turns content into short animations (more in our Claude Motion guide). |
| Where you use it | grok.com, iOS and Android apps, and the Grok tab on X. | Web, desktop, iOS and Android apps. |
| Coding agent | Grok Build: a terminal app that can also run headless or inside other apps, with skills, plugins and MCP servers. Its grok-build-0.1 model is also on the API, and other models can be added in its config file. | Claude Code, included in all paid plans and sharing their usage limits. Its skills follow the Agent Skills open standard. |
| Personal agent | Grok Bot, included in The X Subscription’s Premium, Super and Ultra plans. | Claude Duet, a research preview with a waitlist. |
| Plans | New customers get The X Subscription: Free, Plus, Premium, Super and Ultra, covering Grok, Cursor and X. Existing SuperGrok and X Premium plans continue. | Free, Pro and Max for individuals; Team and Enterprise for organizations. |
| How usage is limited | Paid plans share one weekly usage pool across Grok and Cursor, and Grok Bot draws from the same pool. | Limits reset on a rolling five-hour window, and paid plans add weekly limits. Chat and Claude Code draw from the same pool. |
Prices are left out on purpose. See xAI's pricing and Anthropic's plans page for current numbers. The sources for every row are listed at the end of this page.
What we have run with each
We have not given both the same task. These are separate jobs, listed so you can see what our own data covers:
- Claude: one run of Claude Code with Claude Opus 5.5 made a 30-second narrated explainer on 2026-10-23. The commands, errors and timings are in making a video with Claude Code and HyperFrames.
- Grok: our Grok Bot runs on the same brief included one with Grok Bot's built-in video tool, which says it uses Grok Imagine. It made 15-second clips, and on-screen text came out misspelled twice. That run is in Remotion vs HyperFrames: 12 test videos.
Neither run tells you which assistant is better at video, because the tools and the jobs were different.
How to choose
These questions decide it better than a ranking would.
- Do you need what is happening on X? Grok lists live X integration in the app and an X Search tool on the API. Claude lists general web search.
- Do you need images or video from the same assistant? Grok Imagine is part of the Grok app. Claude's current models output text.
- How much do you need to feed it at once? On the API, the current Claude models list 1M tokens and Grok 4.7 lists 500k.
- Which coding agent fits your setup? Compare Claude Code and Grok Build on your own repository before you commit to a plan.
- What are you already paying for? Paid Claude plans include Claude Code. The X Subscription's paid plans share usage across Grok and Cursor, and Premium and above include Grok Bot.
If you are choosing between the two companies' agents rather than their chat apps, see Grok Bot vs Claude Duet and what Grok Bot is.
Limits of this page
- Everything in the table comes from xAI's and Anthropic's own pages, checked on 2026-10-25. Model names, limits and plans change often.
- We have not run Grok and Claude on the same task, so nothing here says how well either one works. If we do, the results will go on this page with a test date, following how we test.
FAQ
Is Grok or Claude better?
We don't know. We have not run the two on the same job, so we will not name one. Third-party leaderboards publish scores, but they change often. The questions under How to choose are a better guide, and so is a test on your own task.
Which has the bigger context window?
On the APIs, Anthropic lists a 1M-token context window for all four current Claude models. xAI lists 500k tokens for Grok 4.7. Some older Grok models are listed at 1M.
Can Claude search X like Grok?
xAI lists live X integration and real-time web search in the Grok app, and an X Search tool on its API. Anthropic lists web search on every Claude plan. We found no X search feature in Anthropic's docs.
Can Claude make images and videos?
xAI's Grok Imagine makes images and videos in the Grok app and on the API. Anthropic's current Claude models list text as their only output. Anthropic does offer Claude Motion, a beta on Team and Enterprise plans that turns content into short animations.
Grok Build vs Claude Code: what is the difference?
Both are coding agents that run in your terminal. Claude Code is included in paid Claude plans and uses Claude models. Grok Build is xAI's coding agent, powered by its grok-build-0.1 model, and can also be pointed at other models through its config file. We have not compared them.
How much do Grok and Claude cost?
We don't list prices because they change. Anthropic's plans are on claude.com/pricing and xAI's on x.ai/pricing. Both have a free tier and several paid plans with usage limits.
Related on bests.one
- Grok Bot vs Claude Duet — the two companies' personal agents.
- Grok Bot vs Claude Code
- What is Grok Bot
- Can Claude make videos?
- How we test
Sources: Welcome to Grok (xAI docs), Grok product page (xAI), Models (xAI API docs), Grok Build overview (xAI docs), Grok Build 0.1 on API (xAI), The X Subscription pricing (xAI docs), Introducing The X Subscription (xAI), Models overview (Anthropic docs), Claude plans (Anthropic), Get started with Claude Motion (Anthropic help), Claude Duet announcement (Anthropic), Extend Claude with skills (Claude Code docs). Checked on 2026-10-25.