Claude Max vs ChatGPT Pro in 2026: which $100 or $200 plan to buy for coding

Disclosure: Enjoy is our product, and it runs Claude Code and Codex on each lab’s plans. We checked every plan and price below against Anthropic’s and OpenAI’s own pages, and the figures against Artificial Analysis’s and SemiAnalysis’s, on November 1, 2026, and we say where you don’t need Enjoy at all, which is most of the time.

Claude Max vs ChatGPT Pro, compared

Both are where people go when $20 a month stops being enough, and for coding, each pays for its own lab’s agent. “Claude Code is included in all paid plans,” says Claude’s pricing, and OpenAI’s says “ChatGPT Work and Codex are included in your ChatGPT Free, Go, Plus, Pro, Business, Edu, or Enterprise plan.” Each lab sells a $100 plan and a $200 plan. Anthropic renamed its plans on October 30: Max 5x is now Max, and, in its words, “The Max 20x plan is now called the Ultra plan.”

Plan	Best for	Price, checked November 1, 2026	Weekly usage	Five-hour limit	Agent and top models
Claude Max	Claude Code on the hardest work, at $100	$100 a month	5 times Claude Pro	Yes, also 5 times Pro	Claude Code: Opus 5.5; Fable 5.1 up to half the week
ChatGPT Pro 100	Long Codex days, at $100	$100 a month	About 10 times ChatGPT Plus	None	Codex: GPT-6.1 Sol, GPT-6 Astra
Claude Ultra	Long Claude Code runs, or more than a Max week	$200 a month	10 times Claude Pro	None once verified	As Claude Max
ChatGPT Pro 200	Several Codex agents, most days	$200 a month	About 20 times ChatGPT Plus	None	As Pro 100
ChatGPT Pro 500	The most usage, and Ultrafast	$500 a month	About 50 times ChatGPT Plus	None	As Pro 100, with Ultrafast
Two plans, one from each lab	Running out now and then, or a second lab’s review	$40 a month for Claude Pro and ChatGPT Plus; $200 for Max and Pro 100	Each plan’s own	Each plan’s own	Both agents
Enjoy, with any of these	Moving a conversation between accounts and agents	A monthly subscription on top of your plans, free for 7 days (pricing)	Your plans’	Your plans’	Claude Code, Codex and other labs’ agents

Prices are in US dollars, before tax. Anthropic states its multiples. ChatGPT’s are approximate and our arithmetic, because OpenAI’s own pages don’t give them: Pro 100 launched at five times Plus, as VentureBeat reported, OpenAI’s September email put the others at 20 and 25, and its October 29 notes doubled Pro 100 and Pro 500. Each lab counts from its own $20 plan, so 5 times and 10 times here don’t measure the same thing.

Which plan gives more usage

Neither lab’s pages can settle it. Anthropic says of Pro that “The number of messages you can send will vary,” and OpenAI that “Your allowance does not correspond to a fixed number of messages or tasks.” As SemiAnalysis puts it, tiers “are then differentiated by usage multipliers relative only to the provider’s other plans.”

SemiAnalysis is the nearest thing to a common measure. It runs requests, watches how far each plan’s meter moves, and prices what fits at API list prices. On October 5, it found “Limits are quite similar across the board for GPT-6 Astra vs Fable 5.1,” and that at the everyday models, Opus 5.5 against GPT-6.1 Sol, Anthropic was “offering ~5x the API-equivalent value across the board.” Three things qualify that:

• OpenAI has since doubled. SemiAnalysis tested after Pro 200’s cut. On October 29, OpenAI gave Pro 100 and Pro 500 “twice their previous Work and Codex allowance,” and doubled the lower Pro 200 allowance. By our arithmetic, that halves the gap, to about 2.5 times, if Claude’s weeks haven’t changed. SemiAnalysis’s public article hasn’t been updated.
• Claude’s weeks may have changed. Since mid-October, subscribers have measured smaller weeks for the same work, and on October 22, SemiAnalysis said an A/B test was cutting monthly value by about 20% on some accounts. Anthropic’s Boris Cherny said on October 28 that it found “a few smaller things” and no “regression in Claude Code.”
• A dollar of API value buys fewer tasks on Opus. On Artificial Analysis’s comparison, a task costs $11.42 in API prices on Claude Code with Opus 5.5 at max effort, using 14.5 million tokens, and $1.50 on Codex with GPT-6.1 Sol at max effort, using 3.9 million. SemiAnalysis warns that “Token efficiency is also an extremely relevant factor, but the industry unfortunately lacks reliable data here.”

So Claude’s plans have measured more dollars of API value a month, and Codex has spent fewer of them per benchmark task. How many of your tasks a plan holds is something only a week of your own work can show, and both labs show a meter in Settings > Usage.

How the two differ

	Claude Max and Ultra	ChatGPT Pro
Coding Agent Index, best setup	62, Claude Code with Opus 5.5 at max effort	55, Codex with GPT-6 Astra at max effort; 54 with GPT-6.1 Sol
API cost and time per task on the index	$11.42 and 1.0 hour	$7.04 and 24.8 minutes with Astra; $1.50 and 23.7 minutes with Sol
Five-hour limit	On Max; none on verified Ultra	None
What shares the limit	All of Claude, your chats included	Work and Codex; from December, ChatGPT too, apart from Instant chats
Top model’s limit	Fable 5.1 has its own, up to half the week	GPT-6 Astra shares the allowance in Codex
Monthly API credits	$100 or $200; not for Claude Code	$100, $200 or $500; not for Codex through API keys
After the limit	Usage credits, at API rates	Credits, or an instant reset
Billing	Monthly only	Monthly only

The index’s scores come from three benchmarks, FrontierCode v1.1, Terminal-Bench 4.0 and SWE-Atlas-QnA-AA, as Artificial Analysis’s page stood on November 1. Its costs are “pay-per-token API cost, not consumer plan pricing”: on a plan, a task draws on your limits instead. Our Claude Code vs Codex guide compares the two agents themselves.

Claude Max

Best for: the hardest work in Claude Code, at $100, if a five-hour limit doesn’t stop you.

“The Max plan includes five times the Pro plan’s usage every five hours and every week,” says Anthropic’s Max page. Claude Code’s model guide defaults to Opus 5.5 on every paid plan, so what Max adds is time on it, plus “Fable included, up to half your weekly limit,” “Priority access at high traffic times” and “$100 a month in API credits to build on the Claude Platform.”

Price: $100 a month. “The Max plan is currently available as a monthly subscription only,” and “Mobile pricing may vary depending on your app platform.”

Watch for: three limits. “Max plans have a five-hour limit, a weekly usage limit that applies across all models, and a separate weekly limit for Fable.” “These limits are shared across Claude and Claude Code, meaning all activity in both tools counts against the same limits,” so long chats leave less for coding. The API credits don’t help there: Anthropic’s docs list “Claude Code, extra usage in the Claude apps” under what they don’t cover. And Anthropic reserves the right to go further: “we may limit your usage in other ways, such as weekly and monthly caps or model and feature usage, at our discretion.”

ChatGPT Pro 100

Best for: long Codex days at $100, with no five-hour stop and a low cost per task.

OpenAI’s usage guide says “Pro 100, Pro 200 and Pro 500 currently have no five-hour usage limit in Work and Codex,” so you can spend the week’s allowance on one long day. Its pricing lists “Longer Codex and ChatGPT Work sessions,” “Pro reasoning powered by GPT-6 Astra” and “Dot, your always-on agent.” Since October 29, Pro 100 has twice its earlier allowance, which is why it’s now about 10 times Plus for five times the price.

Price: $100 a month. “Pro is billed monthly. Annual billing and payment for multiple months in advance are not available.”

Watch for: the allowance is about to change shape. OpenAI’s help center says “Starting in December, Plus, Pro, and Business plans’ usage allowance will be shared across ChatGPT and Codex,” measured in tokens, and that “Instant chat will not draw from your usage.” It hasn’t named the day or the new size, as our memo explains. “A larger allowance does not guarantee twice as many completed tasks,” and Fast mode “uses your included allowance at 2.5 times the rate.” The API credits don’t buy Codex: “This benefit does not cover Codex usage through API keys.” And the plans move: OpenAI announced a cut to Pro 200 on September 29 and canceled it a month later.

Claude Ultra

Best for: long Claude Code runs that a five-hour limit would interrupt, or needing more than a Max week.

“Ultra includes 10 times the Pro plan’s usage every week and has no five-hour limit, for verified accounts,” says Anthropic’s Ultra page, and “Existing Max 20x subscribers move to Ultra automatically, with no price change.” The old name only ever fit the five-hour limit: “When we added weekly limits to Max 20x, we set them to 10x Pro’s (the 5-hour limit stayed at 20x),” wrote Mike Overell, who leads product for Claude subscriptions. It has everything in Max, with “$200 a month in API credits, up from $100 on Max.”

Price: $200 a month. “The Ultra plan is currently available as a monthly subscription only.”

Watch for: it costs twice Max for twice Max’s week, so the extra $100 buys the second week’s worth and the end of the five-hour stop. That end waits on verification: “If your account isn’t verified yet, a temporary limit applies,” until Anthropic verifies it “using factors like your account and payment history,” and its page gives no time for that. Fable keeps “a separate weekly limit” here too.

ChatGPT Pro 200

Best for: several Codex agents running most days.

Pro 200 has the same features as Pro 100, at twice the price and about twice the allowance. “Pro 200 includes more usage than Pro 100,” says OpenAI’s Pro article, and the September cut is gone: “If you had the lower allowance, it has doubled to match.” OpenAI’s pages name no verification step for its lack of a five-hour limit, as Anthropic’s do for Ultra’s.

Price: $200 a month, with $200 a month of API credits.

Watch for: “Your subscription remains Pro 200. Ultrafast is available only on Pro 500.” By the figures above, Pro 100 and Pro 200 cost the same per Plus-sized allowance, so move up only if a Pro 100 week runs out.

ChatGPT Pro 500

Best for: the most usage either lab sells to one person, and Ultrafast.

“Pro 500 offers the highest included usage of the three plans,” and Claude has no plan for individuals above $200. It’s about 50 times Plus by the same arithmetic, and it adds Ultrafast for GPT-6 Astra and GPT-6.1 Sol.

Price: $500 a month, with $500 a month of API credits.

Watch for: Ultrafast “uses your plan’s included usage first at 8× the Standard rate,” so an Ultrafast week is an eighth as long. If you’d buy it only for the usage, see first whether Pro 200 runs out.

Two plans, one from each lab

Best for: running out now and then, or wanting a second lab’s review of the same code.

Anthropic’s and OpenAI’s limits are separate, so when one agent runs out, the other still works. Claude Pro and ChatGPT Plus together cost less than half of either $100 plan, and for the $200 of one top plan you can have Max and Pro 100, half the size at each lab. One file can instruct both: Claude Code “can read AGENTS.md as your project instructions,” per its docs, and our guide to using both covers the rest.

Price: $40 a month for Claude Pro and ChatGPT Plus, or $37 with Claude Pro billed yearly. $200 for Max and Pro 100.

Watch for: the $20 plans each have a five-hour limit. OpenAI describes Plus as “Power a few focused coding sessions each week,” with an estimate of 15 to 160 local messages every five hours on GPT-6.1 Sol, and “These estimates are not fixed message limits.” On Claude Pro, Fable is “paid with usage credits.” And each agent keeps its own conversations, so you carry the context across unless a tool does it for you.

Enjoy, with any of these

Best for: paying for both labs, and moving a conversation from one to the other when a plan runs out.

Enjoy isn’t a Claude or ChatGPT plan: it’s a desktop app for Mac, Windows and Linux that runs Claude Code and Codex, with other labs’ agents and local models through OpenCode, on the plans you already have. What it adds to this choice:

• What’s left, side by side. Its usage meters show each signed-in account’s limits and when each resets, so you can run the week of your own work this guide recommends, on both plans.
• One conversation, either agent. Switch agents at any point: “The new agent gets the conversation so far, including replies from other agents, before it works on your message.”
• Limits handled. When an agent runs out, Enjoy switches to another of your accounts with usage left, or waits for the reset and carries on.
• It tells you when you’re overpaying. On Claude Max, it says when your last four weeks would have fit a smaller plan. It never changes your plan.

Price: Enjoy is one monthly subscription, and the first seven days are free, with a card to start. It includes your phone and any browser and your other computers, and each teammate you invite needs a subscription of their own. See pricing.

Watch for: Enjoy doesn’t change either lab’s limits, or add Claude to a ChatGPT plan or GPT to a Claude one. Its plan suggestions cover Claude Code, not ChatGPT, and still use the names Max 5x and Max 20x. Enjoy runs Claude Code headless, so if the reports in issue #97074, still open, are right, Claude Code in Enjoy uses the five-hour window faster than in a terminal. That matters on Max, and not on a verified Ultra account, which has no such window. Waiting for a reset needs “Your computer must be on, with Enjoy open, when the usage resets,” and scheduled workflows run only while it’s awake with Enjoy open. Enjoy isn’t open source, and it’s a monthly subscription on top of your AI plans once the free week ends.

When to stay on the plan you have

If you’re on a $20 plan and run out once or twice a month, stay. Both labs sell the odd afternoon by the token, usage credits on Claude and credits on ChatGPT, and our guides to each lab’s plans, Claude’s and ChatGPT’s, work out when an upgrade pays. Upgrade when it happens most weeks, not after one bad one.

If you already pay for one of these and its agent does your work, the other lab’s numbers aren’t a reason to move: an agent you trust is worth more than a bigger meter. And don’t move over December’s change yet. OpenAI says “We expect most users won’t need to wait for their usage to reset more often than they do now,” and it hasn’t published the allowance that would prove it. Every plan here bills monthly, so you can try the other lab for a month when you have a reason to.

How we compared

We read Anthropic’s pricing, its help center articles on Pro, Max, Ultra, Claude Code and usage credits, and its docs on models, instructions and API credits; OpenAI’s Codex and ChatGPT pricing, its help center articles on Pro, usage, credits, API credits and the December change, and its release notes; Artificial Analysis’s Claude Code vs Codex comparison; and the public part of SemiAnalysis’s October 5 article, its October 22 post and Overell’s and Cherny’s posts, all on November 1, 2026, and every one is linked below. We didn’t measure either plan ourselves. SemiAnalysis’s figures are for its own mix of agentic work, and the ChatGPT multiples and the halved gap are our arithmetic, labeled as such. We quote each company’s own words for its limits and leave out what it hasn’t published, including message counts for Claude and the size of December’s allowance. Plans and limits change often, so we recheck this guide against its sources at least once a month, and the date at the foot of the page says when we last did.