The line-up
OpenAI's pricing splits into three groups, and the confusion usually comes from mixing them up:
- Consumer ChatGPT: Free, Go, Plus and Pro. One person, paid monthly by card. Pro comes in two usage tiers — OpenAI's page describes them as 5× and 20× the Plus allowance — at two prices.
- Business ChatGPT: Business (per seat, two-seat minimum, monthly or annual) and Enterprise (custom annual contract, invoice billing). Also Edu for universities and a free Teachers plan for US K-12 educators.
- The API: metered per token, no seats, priced per model on OpenAI's API pricing page. This is what developers and products pay; it has nothing to do with a ChatGPT subscription.
The table below lists the ChatGPT tiers with verified prices. The API is not in it because it has no monthly price — see the last section.
| Plan | Price | Model Access | Limits |
|---|---|---|---|
| ChatGPT Free | Free | GPT-5.2 Instant, GPT-5.2 Mini (fallback) | Dynamic message and tool caps — limited usage envelope; credits can unlock extr… |
| ChatGPT Go | Monthly: $6.00/mo | GPT-5.2 Instant, GPT-5.2 Mini (fallback) | Expanded message, upload, and image caps over Free; flexible credits extend sel… |
| ChatGPT Plus | Monthly: $20.00/mo | GPT-5.2 Instant, GPT-5.2 Thinking, GPT-5.2 Mini (fallback), Legacy models (plan… | High personal limits across messages, uploads, image generation, deep research,… |
| ChatGPT Pro | Monthly: $200.00/mo | GPT-5.2 Instant, GPT-5.2 Thinking, GPT-5.2 Pro, GPT-5.2 Mini (fallback), Legacy… | Highest individual limits across chat, deep research, agent mode, image generat… |
| ChatGPT Business | Monthly: $30.00/mo | GPT-5.2 Instant, GPT-5.2 Thinking, GPT-5.2 Mini (fallback), Legacy models (plan… | Per-seat usage envelope with monthly/annual billing and flexible pricing overag… |
| ChatGPT Enterprise | Monthly: Contact sales | GPT-5.2 Instant, GPT-5.2 Thinking, GPT-5.2 Pro, GPT-5.2 Mini (fallback), Legacy… | Contract-defined usage envelope with enterprise governance controls |
What each step up buys
Reading OpenAI's own feature grid top to bottom, each tier adds one thing that the previous one caps:
- Free → Go: room. More messages with tools, more uploads, more image generation, a doubled context window for the fast model, longer memory. Go may show ads. See the ChatGPT Go guide.
- Go → Plus: capability. The reasoning models, a much larger reasoning context, projects, scheduled tasks, custom GPTs, and a Codex allowance large enough to code with. This is the step most people should stop at.
- Plus → Pro: volume. Five or twenty times the usage, the Pro-tier reasoning model, and unlimited image generation. Pro exists for people who hit Plus's walls daily, not for people who want "the best one".
- Plus → Business: administration, not capability. A Business seat is a Plus seat with a workspace, SSO and unified billing around it. See the Business pricing guide.
- Business → Enterprise: governance and uncapped usage. SCIM, key management, data residency, audit APIs, and no fixed rate limits. See the Enterprise pricing guide.
The number that changes most between tiers
OpenAI's pricing page publishes a context figure per tier, and it moves more than the price does. For the fast model, the total context is 27,000 tokens on Free, 54,000 on Go and Plus, and 128,000 on Pro. For the reasoning models it is listed as varying on Free, then 256,000 on Go and Plus and 400,000 on Pro.
OpenAI also notes that the space available for your own input is a fraction of the total, because system instructions, memory and the model's own reasoning share the same window. If you work with long documents or large codebases, that footnote matters more than the headline. The context window guide explains how to read these numbers across vendors.
The API: a different product with a different meter
OpenAI's API is billed per million tokens, with a separate input and output price for each model, listed on OpenAI's API pricing page. There is no subscription, no seat and no allowance — you pay for what you send and receive. A ChatGPT Plus subscription does not include API credit, and API spend does not give you ChatGPT.
The API is the right meter when the usage is a program rather than a person: an integration, an automation, a product feature, or Codex running unattended. It is usually the wrong meter for one person who chats all day, because a flat subscription caps the cost. The API vs subscription guide has the break-even arithmetic, and the Codex pricing guide covers the one product that can be paid for either way.
Which one, in one line each
- Trying it: Free.
- Free keeps stopping, you just chat: Go.
- You want the reasoning models or Codex: Plus.
- You exhaust Plus most days: Pro, starting with the 5× tier.
- A team needs one bill and SSO: Business.
- IT needs SCIM or data residency: Enterprise.
- The user is software, not a person: the API.
And if the real question is OpenAI versus someone else at the same price, the Claude Pro vs ChatGPT Plus comparison is where most people start.
Frequently Asked Questions
No. ChatGPT plans and the API are separate products with separate billing. The API is metered per token on OpenAI's platform; a Plus, Pro or Business subscription does not include API credit.
Usage. OpenAI sells Pro in a 5× tier and a 20× tier, the multiplier being relative to the Plus allowance. Both get the Pro-tier reasoning model and unlimited image generation; the higher tier is for people who exhaust the lower one.
Per seat, Business is priced close to Plus, with a lower annual rate and a two-seat minimum. You are paying for the workspace — one bill, SSO, an admin console, shared GPTs — not for extra capability per person, which is the same as Plus.
Nonprofits can get up to 75% off Business and Enterprise through OpenAI for Nonprofits. Business has a lower annual rate. US K-12 educators can use ChatGPT for Teachers free until June 2027. There is no published student discount on the consumer plans.