ChatGPT Pricing: Every Plan Compared

ChatGPT pricing runs from $0 to $200 a month for individuals, then moves to per-seat pricing for teams. There are six named tiers: Free, Go, Plus, Pro, Business, and Enterprise. Each one gates model access, message volume, and Agent mode differently.

The gap that trips people up is not the sticker price. It is the message cap. Plus looks generous until a heavy work week hits the 3-hour rolling limit on the flagship model, and the account quietly falls back to a lighter model mid-conversation.

This page lays out what you get at each price point, who actually needs Pro versus Business, and the seat mechanics OpenAI does not put on the pricing page. Prices and limits change often, so treat the numbers below as a snapshot and verify current figures on OpenAI's own pricing page before you buy.

What Are the ChatGPT Pricing Tiers

ChatGPT has six pricing tiers: Free, Go, Plus, Pro, Business, and Enterprise. Free and Go serve casual individual use, Plus and Pro serve power users, and Business and Enterprise serve organizations that need admin controls and data protections.

OpenAI renamed the old Team plan to ChatGPT Business in 2025 as a naming change only, then cut Business seat pricing in April 2026. Feature parity between the renamed plan and the old Team plan stayed the same at launch, and pricing has moved since.

Each tier layers on top of the one below it. Plus adds more volume and features over Free and Go. Pro removes most caps. Business and Enterprise add seat management, admin controls, and data-retention guarantees on top of Plus-or-better model access.

• Free: $0, ad-supported, capped access to the flagship model
• Go: $8/month, more volume than Free, still ad-supported, no advanced reasoning or Agent mode
• Plus: $20/month, higher caps, Agent mode, image and voice tools
• Pro: $200/month, effectively unlimited flagship access, priority compute
• Business: $20-25/seat/month depending on billing term, admin console, no training on business data by default
• Enterprise: custom quote, typically requiring a seat minimum and annual contract

ChatGPT Free and Go: What You Get at the Bottom

Free and Go are OpenAI's entry tiers, and both now carry ads as of OpenAI's February 2026 rollout of advertising. Free gives every signed-up user a capped number of messages on the flagship reasoning model before falling back to a lighter model for the rest of the reset window.

Go costs $8 a month and raises the message ceiling, upload limits, image generation volume, and memory length compared to Free. It still excludes Agent mode, Deep Research, Codex, and the top-tier reasoning models that Plus and above unlock.

Go launched in India in 2025 as a lower-cost regional tier, then rolled out globally in January 2026. It fills the gap between a free account and the $20 Plus commitment, but it is not a business tool. There is no admin console, no seat pooling, and no data-retention guarantee.

ChatGPT Plus vs Pro: Where the $180 Gap Goes

ChatGPT Plus costs $20/month and Pro costs $200/month, and the difference is volume plus priority, not a different product. Plus gives ad-free access to the flagship model with a rolling message cap that resets every few hours, plus a separate weekly cap on extended-reasoning messages.

Pro removes most of that ceiling. OpenAI describes Pro access as unlimited subject to abuse guardrails, meaning there is no published cap for normal human use, only automated-abuse protections. Pro also gets priority access during high-traffic periods, so responses stay fast when Plus users see slowdowns.

Both tiers include Agent mode, image generation, and voice. The practical decision point is usage pattern: if a Plus account regularly hits its reasoning-message cap mid-week and falls back to a lighter model, Pro pays for itself in unblocked work hours rather than in extra features.

• Plus: $20/month, rolling message caps, weekly extended-reasoning cap
• Pro: $200/month, no published cap for normal use, priority compute
• Both: Agent mode, image tools, voice, ad-free

ChatGPT Business Pricing: Per-Seat Cost and Minimums

ChatGPT Business costs $20 per seat per month billed annually, or $25 per seat per month billed monthly, after an April 2026 price cut from the prior $25/$30 rates. A ChatGPT Business workspace requires a minimum of 2 seats to purchase.

Business is the plan built for small and mid-size teams that need a shared billing account, an admin console, and a guarantee that conversation data is not used to train OpenAI's models by default. It includes the same model access as Plus, now including unlimited flagship-model use per OpenAI's 2026 update.

What Business does not include is SSO enforcement at scale, dedicated data residency, or the deepest compliance controls large enterprises need. Those live in Enterprise. Teams under roughly 20-30 seats usually land on Business; larger orgs with compliance requirements typically get pushed to an Enterprise quote regardless of headcount.

• Annual billing: $20/seat/month
• Monthly billing: $25/seat/month
• Minimum: 2 seats
• Includes: admin console, workspace billing, no default training on business data

ChatGPT Enterprise Pricing: Why There Is No Sticker Price

ChatGPT Enterprise has no published price and requires a custom quote from OpenAI's sales team. Market reporting on 2026 deals converges on roughly $45 to $75 per seat per month, with most negotiated contracts landing near $50-$60 at seat counts of 150 or more.

OpenAI does not publish an official seat minimum, but procurement reports consistently cite around 150 seats as the practical floor sales teams quote against, and Enterprise contracts run on annual, prepaid terms rather than month-to-month billing.

Enterprise adds SSO/SCIM provisioning, dedicated data residency options, longer context windows on supported models, and the compliance paperwork (SOC 2 reports, DPAs) that procurement teams require. Verify current per-seat ranges and minimums directly with OpenAI sales, since none of this is posted publicly and quotes shift by volume and term length.

Which ChatGPT Plans Include Agent Mode

Agent mode is available on Plus, Pro, Business, and Enterprise, but not on Free or Go. OpenAI's help center confirms agent access starts at the Plus tier and carries through every paid plan above it, including Edu accounts.

That makes Agent mode a hard line between the two casual tiers and everything else. A team evaluating ChatGPT for browsing tasks, form-filling, or multi-step web actions needs at least a Plus seat per user, or a Business/Enterprise workspace, since Free and the $8 Go plan cannot run it at all.

Message volume for Agent mode tasks draws from the same caps as regular flagship-model messages on Plus, which is a common surprise: an agent run that takes several minutes can burn through a meaningful chunk of a Plus account's rolling cap in one session.

• Free: no Agent mode
• Go ($8/mo): no Agent mode
• Plus ($20/mo): Agent mode included, shares the Plus message cap
• Pro ($200/mo): Agent mode included, effectively uncapped
• Business / Enterprise: Agent mode included org-wide

How ChatGPT Message Caps Actually Work

ChatGPT message caps reset on a rolling window rather than a fixed daily quota, and they differ by model inside the same plan. Free grants a small number of flagship-model messages roughly every few hours, then automatically routes the rest of a session to a lighter, faster model until the window resets.

Plus raises that ceiling and adds a separate weekly allowance for extended-reasoning messages, which are the slower, more thorough model runs used for complex problems. Once that weekly reasoning allowance is gone, the account can still chat, just without the deeper reasoning mode until the week rolls over.

This is the part generic pricing pages skip: the fallback is silent. A Plus user mid-task will not get an error, they will just get shallower answers from a smaller model once a cap hits, which can look like a quality regression rather than a billing limit. If you are debugging why ChatGPT output quality dropped mid-session, check whether a cap was hit before assuming the model changed on OpenAI's end.

How to Pick the Right ChatGPT Tier

Match the tier to usage pattern and headcount, not to feature lists alone. A solo user who chats occasionally is fine on Free or Go. A solo power user who hits caps weekly should go Plus first and only jump to Pro once Plus caps become a recurring blocker.

For teams, the decision is seat count and compliance need more than raw usage. Under roughly 20-30 seats with no strict data-residency requirement, Business at $20-25/seat is the simpler and cheaper path. Past that scale, or with SSO/SCIM and residency requirements, an Enterprise quote is usually unavoidable regardless of seat count.

When we run monetization and tooling audits across the routine automation portfolio we manage for clients, the recurring pattern is that teams over-provision Pro seats for people who never hit Plus caps, and under-provision Business seats for people who do. Auditing actual per-user message volume before renewal, not guessing from job title, is what catches both mistakes.

• Occasional individual use: Free or Go
• Frequent individual use, hits Plus caps: Pro
• Small team, no strict compliance need, under ~30 seats: Business
• Large org, needs SSO/SCIM, data residency, or 150+ seats: Enterprise (custom quote)

How ChatGPT Pricing Compares to Claude and Gemini

ChatGPT's per-seat pricing sits close to Claude's Team and Enterprise tiers and Gemini's Business tier, with the real differentiator being caps and included tools rather than sticker price. All three vendors run the same basic pattern: a free tier, a $20-ish individual paid tier, and custom enterprise pricing above that.

The comparison that matters for a buying decision is not the monthly fee, it is which specific capabilities (Agent mode, deep research, code execution, context window) are gated behind which tier, since that gate line moves independently for each vendor and changes with every model release.