Claude Pro vs ChatGPT Plus: 1M tokens of context against unlimited everyday chat, both $20

Which is better, Claude Pro or ChatGPT Plus?

On 18 October 2026 the Artificial Analysis leaderboard ranked Claude Opus 5.5 first of every model it tests, at 58 on its Intelligence Index. Claude Sonnet 5.5 came second at 56. Both of them are in Claude Pro's chat, and neither costs anything extra. The best model on ChatGPT Plus is GPT-6 Astra at 53, and OpenAI lets Plus use it only in ChatGPT Work and Codex. In ordinary chat, Plus runs GPT-6 Sol. It scores 48.

Those are each model's best runs, at max effort. OpenAI's model guide gives Plus the Medium and High thinking levels in chat, and not Extra High. So Plus chat runs GPT-6 Sol below its top setting. Claude Pro has no such cut: Anthropic's effort settings article gives Claude Opus 5.5 an effort selector that goes up to Max, and the only plan rule it names is that Enterprise admins can switch levels off. Claude Fable 5.1 also scores 53, but Anthropic bills Pro users for it with usage credits, outside the plan.

What Plus gives back is room. OpenAI's pricing page lists everyday text chat on Plus as "Unlimited", with an asterisk for abuse guardrails. Anthropic meters every Claude Pro message against a five hour session limit and a weekly limit, and it doesn't publish either number. Plus also makes images. Claude doesn't.

What do Claude Pro and ChatGPT Plus include for $20?

Both plans cost $20 a month in the US, and both include memory, projects, file uploads, voice, a research mode and a coding agent. They part on the rows below. We read each company's plan page and help center on 18 October 2026 for this table.

Context is the hard number in that table. Anthropic's usage and length limits article says that on paid plans "the newest models support up to a 1M token context window, while others support 500K or 200K tokens." OpenAI's plan table gives Plus 256K on its reasoning models and 54K on Instant, which it puts at about 320 and 40 pages of text. So Claude Pro holds about four times as much of a long contract or codebase in one chat.

Images are the one feature Claude Pro lacks outright. Anthropic's help center says "Claude doesn't generate photos or illustrations the way image-generation tools do." What it can do is build charts, diagrams and interactive visuals. ChatGPT Plus has Images 2.5, which OpenAI shipped on 8 September 2026 according to its release notes.

Which has higher usage limits, Claude Pro or ChatGPT Plus?

ChatGPT Plus has the higher limit for everyday chat, and neither company prints a message count for its strongest chat model. OpenAI lists Plus chat as unlimited. The Medium and High thinking levels have limits of their own that OpenAI doesn't number, and when you reach one, its model guide says "ChatGPT may continue with another available model." You aren't locked out. You get a smaller model.

Claude Pro works the other way. Anthropic's help center describes "a session limit that resets every five hours, and weekly limits that reset once every week." Claude Code spends the same pool. The Claude Code with Pro article says "all activity in both tools counts against the same limits." When you hit the wall, you wait, upgrade, or keep going on usage credits that you pay for.

OpenAI does print numbers for ChatGPT Work and Codex, the agent side of Plus, and it keeps that allowance apart from Chat. Its Work and Codex usage guide, updated 10 October 2026, gives these ranges for one five hour window on Plus. It adds that "these ranges are estimates, not guaranteed message counts," and that a weekly limit may apply on top.

Plus has smaller caps on some of the other things you do with it, too. OpenAI allows up to 80 file uploads every 3 hours, per its file upload article, and 3 hours of voice in a rolling 24 hours, per its Voice article. Anthropic prints no hourly or daily cap on either for Pro. Its upload article allows up to 20 files in one chat. Is Claude Pro worth it covers how fast heavy users reach the Claude wall.

Is Claude Pro or ChatGPT Plus cheaper in your country?

ChatGPT Plus costs the same as Claude Pro, or less, in 16 of the 18 App Store countries we read on 18 October 2026. Claude Pro costs 9% to 30% more in six of them. Brazil has the widest gap, at R$129.90 a month against R$99.90. That's R$360 more over a year. Claude Pro is cheaper only in Germany and Nigeria, and France's store, which we also checked, shows the same euro prices as Germany's.

We opened the Claude and ChatGPT listings in each country's App Store on 18 October 2026 and compared the monthly in app price in local money. No exchange rate sits in the table, so each row holds for that store alone. We didn't read Google Play.

On the web, a year tips it the other way. Claude Pro is $20 a month on claude.com, or $200 for a year, which is $17 a month. OpenAI's Plus help article says: "Currently, we do not support annual billing or the option to pay for multiple months in advance for ChatGPT Plus subscriptions." So a year of Plus costs $240 on chatgpt.com, $40 more than a year of Claude Pro. Apple charges $214.99 for the Claude year, which is $14.99 more than claude.com. Apple's US store also lists a second ChatGPT Plus item at $200, and the listing doesn't say what term it covers.

Anthropic adds that its "Prices shown don't include applicable tax." For the full country lists, with the dollar value of each price, see Claude Pro price by country and ChatGPT Plus price by country.

Is Claude Pro or ChatGPT Plus better for coding?

Claude Pro is better for coding on the scores, and ChatGPT Plus gives coding an allowance of its own. On Terminal-Bench 4.0, Artificial Analysis's test of an agent at work in a terminal, Claude Sonnet 5.5 scores 64% and Claude Opus 5.5 scores 60%. GPT-6 Astra scores 59%, GPT-6.1 Sol 56% and GPT-6 Sol 44%, read 18 October 2026. On SciCode, which asks for scientific code from a written spec, Opus leads Astra by 67% to 56%.

The plans split on how coding is paid for. Claude Code comes with Pro and draws on the same session and weekly limits as chat. A long refactor can leave you no chat until the reset. Codex on Plus draws on the Work and Codex allowance, which OpenAI keeps apart from Chat. The catch is in how much Astra it gives you. OpenAI estimates 5 to 45 Astra messages per five hours on Plus, and says Plus seats "include limited Astra usage, with optional credits for additional usage afterward."

So pick Claude Pro if your code needs the strongest model. Pick Plus if you code and chat all day and can work on GPT-6.1 Sol, which OpenAI estimates at 15 to 160 messages per window.

Should you pick Claude Pro or ChatGPT Plus?

Pick Claude Pro for hard problems and long files, and ChatGPT Plus for many messages, images and voice. Each line in the table has its number from the sections above, read 18 October 2026.

Claude Pro loses when you need volume you can plan around. Anthropic prints no message count, a session can end in the middle of a task, and Claude Code spends the same pool as chat. And it can't make an image for you at all.

ChatGPT Plus loses on hard problems and long files. Its chat model scores 48 at max effort, and Plus can't pick Extra High, the top level OpenAI offers in chat. Its strongest model, Astra, is capped at an estimated 5 to 45 messages per five hours and is missing from chat entirely. And 256K tokens is about a quarter of what Claude reads. If you already pay for one and wonder about the other, is ChatGPT Plus worth it and do you need both take that question on.

What if you want Claude and GPT on one plan?

Both $20 plans together cost $40 a month. Whizi Pro is $29.99 a month for 2,500 credits as of October 2026, with Claude Opus at 20 credits a message and GPT-6 Sol at 10. That buys 125 Opus messages or 250 GPT-6 Sol messages a month, and side by side chats send one prompt to both. Whizi loses on volume and tools, though. Its Opus cap is printed and small, it has no Claude Code or Codex, and voice is 80 minutes a month. If you live in one lab's app all day, that lab's own $20 plan is the better buy. Whizi vs Claude Pro compares the two plans.