Grok vs Claude (2026): Prices, Limits, Coding Tools and Which to Pay For

The short answer

As of October 23, 2026, Claude Pro is $20 a month ($17 billed annually) and includes Claude Code. Grok's recommended tier, Premium, is $30 and bundles the Cursor code editor and X Premium. Grok also sells an $8 Plus tier, which Claude has no equivalent of, and both go up to $200 for individuals. For a team, Claude Team is cheaper per seat ($20 billed annually) than Grok Teams (from $30). The catch most comparisons miss: if Claude is the model you want to code with, Grok's Cursor bundle doesn't include it. Claude models in Cursor are billed on top at API prices.

Grok's prices changed on October 12, when x.ai replaced the SuperGrok plans with The X Subscription, one plan that covers Grok, Cursor and X. A lot of Grok-vs-Claude tables online still show SuperGrok. Every number below comes from x.ai's and Anthropic's own pricing pages, read today.

Grok vs Claude plans, side by side

x.ai recommends Premium for new AI users and Plus "for social creators with limited AI usage," so the $8 tier is mostly an X Premium subscription with some Grok attached. The full Grok ladder, old prices included, is on SuperGrok pricing, and every Claude plan is on Claude AI pricing.

The coding catch: Cursor in Grok vs Claude Code in Claude

Both $20-to-$30 plans now come with a coding tool, and this is where the two really split.

• Grok Premium ($30) includes Cursor, the AI code editor, on the same usage pool as your Grok chats. Inside that pool you get Grok's own models. x.ai's page says third-party models in Cursor, which includes Claude, are "billed separately" at API list prices. On the $8 Plus tier, Cursor runs Grok models only.
• Claude Pro ($20) includes Claude Code, Anthropic's coding agent, and Anthropic's page says it "shares the same usage limits as the rest of your plan." You get Claude's models for coding with no second bill, until you hit the limit.

So if you picked Grok for the Cursor bundle but the model you actually want writing your code is Claude, you pay $30 for Grok plus Claude's API rates on top. That's usually more than Claude Pro, or Max if you code all day. If you're happy coding with Grok's models, the bundle is a real deal: x.ai priced the three products bought separately at $90 a month when it launched The X Subscription. What Claude Code costs at each tier is on Claude Code pricing, and Cursor's own plans are on Cursor pricing.

How each one meters usage

Grok: one usage pool, shared between Grok (chat, Imagine, Voice and Grok Bot) and Cursor, that refreshes weekly. x.ai states the tiers as multiples of Plus (Premium 4x, Super 15x, Ultra 50x) without saying what Plus itself buys, so there's no base number to plan against. Run out and you buy top-ups or wait. Grok API access comes with paid plans, but API calls are billed separately. The missing numbers are covered in Grok usage limits.

Claude: limits reset on a 5-hour session, with weekly limits on top for paid plans. Max is defined against Pro, at 5x or 20x Pro's usage per session, and Claude Code, chats and Cowork tasks all draw from that one pool. Past a limit you can wait, upgrade, or turn on usage credits billed at standard API rates. Plan by plan, that's on Claude usage limits.

Neither one gives you a fixed message count. The practical difference is the clock: Claude's limit can stop you mid-afternoon and come back in a few hours, while Grok's weekly pool lets one heavy day eat into the rest of your week.

What a team pays: Grok Teams vs Claude Team

Those are standard seats. On Claude Team you can give the one or two heavy users a premium seat ($100 billed annually, five times a standard seat's usage) instead of upgrading everyone. Grok Teams also mixes seat types, at 4x to 50x Plus depending on the seat, and includes Cursor in every seat, which matters if half the team writes code. Invoice billing, SCIM and audit logs are Enterprise-only on Grok. Claude Team puts SSO and central billing in the base plan. The seat-by-seat detail is on Claude Team plan.

Privacy: the setting to check before you paste in client data

On Claude's individual plans (Free, Pro and Max), model training on your chats is a setting you turn off. On Team and Enterprise, Anthropic doesn't train on your content by default. x.ai's pricing page doesn't say what the default is on its individual plans. Grok Teams lists team-wide privacy settings and custom chat data retention. Whichever you buy, an owner putting client or company data into a personal plan should check that setting on day one, not after.

Which one should you pay for?

• You post on X and code in Cursor with Grok's models: Grok Premium at $30. It replaces subscriptions you'd buy anyway.
• You want Claude to write your code: Claude Pro at $20, or Max at $100 if you'd hit Pro's limit by lunch. Weigh that on is Claude Max worth it.
• You want one assistant for writing, research and documents: Claude Pro at $20 is $10 a month less than Grok Premium. Whether Grok's assistant alone is worth the extra is what is Grok worth it covers.
• You're buying for a team of 5 to 25: Claude Team has the lower seat price and the no-training default, unless most of the team codes in Cursor every day.
• You're also weighing ChatGPT: the same comparison against OpenAI's plans is on Grok vs ChatGPT.

Neither vendor publishes a head-to-head on answer quality, and both ship new models every few weeks, so I won't give you a "which is smarter" score that's out of date by November. Both have free tiers. Run your five most common real tasks through both on the same day, then pick from the prices above.

For a team, the subscription is the cheap part. The real cost is deciding what the tool may touch, what people use it for, and how you'd know it's working. When a company wants that set up properly around the plan it picked, that's the AI implementation work I do, at a published price.

Sources

Read on October 23, 2026: x.ai/pricing (The X Subscription tiers, Teams and Enterprise features, the usage-pool FAQ, API billing, third-party models in Cursor billed separately) and claude.com/pricing (Free, Pro $20 or $17 annual, Max 5x and 20x, Team standard and premium seats for 2 to 150, Enterprise $20 a seat plus usage with a 20-seat minimum, Claude Code included on paid plans and sharing their limits, usage credits at API rates). x.ai's launch post for The X Subscription (October 12, 2026; read October 15) for the $90 separate-purchase figure. Training defaults by plan as published on claude.com/pricing and Anthropic's Privacy Center. Team totals are my arithmetic from the published per-seat prices; prices in USD, before tax.