AI cost · OpenAI
ChatGPT cost: what you actually pay
ChatGPT Business is a flat per-seat price, so the value question is utilization: orgs buy $30 seats that go idle, put light users on $200 Pro, or jump to Enterprise (≈150-seat minimum) before they need it.
Public pricing
| Plan | Price | Billing | Includes |
|---|---|---|---|
| Free | Free | monthly | — |
| Plus | $20 / mo | monthly | — |
| Pro | $200 / mo | monthly | — |
| Business | $30 / user/mo$25/user/mo on annual billing; minimum 2 users. (Formerly "Team".) | monthly or annual | — |
| Enterprise | CustomCustom pricing; typically ~150-seat minimum. | custom | — |
Last updated: 2026-07-22. Sourced from ChatGPT pricing (official) ↗, ChatGPT plans comparison (2026) ↗.
Where the real ChatGPT cost hides
The sticker is the smallest part of the story. For ChatGPT, these are the line items that quietly inflate the bill:
- Idle per-seat Business/Enterprise licenses
- Light users parked on $200 Pro who would fit Business
- Jumping to Enterprise minimums before utilization justifies it
What an audit finds on ChatGPT
Business seats provisioned org-wide but actively used by a fraction of the roster, plus a few $200 Pro seats that could drop to Business.
ChatGPT pricing FAQ
How much is ChatGPT Business?
$30/user/mo, or $25/user/mo on annual billing, with a 2-user minimum.
What is the difference between Business and Enterprise?
Business is self-serve from 2 users; Enterprise adds faster/unlimited access, larger context, SLAs, data residency, custom terms, and typically a ~150-seat minimum.
Does ChatGPT bill inactive seats?
Yes — Business and Enterprise are per seat, so unused assigned seats keep billing until removed.
What is ChatGPT really costing you?
Spendassay connects your ChatGPT spend to actual usage and engineering output — a proof-level AI cost report with the wasted dollars named. Neutral across every AI vendor.
Free · read-only · no card
Other chat & enterprise assistants pricing
More: the complete AI cost management guide · every tool, with prices · cost per outcome · compare platforms · how we measure