OpenAI ChatGPT and API licensing

OpenAI ChatGPT and API licensing is how OpenAI sells access to its models to organizations and individuals. Business customers contract under the OpenAI Services Agreement, which “only applies to use of OpenAI’s APIs, ChatGPT Enterprise, ChatGPT Business, ChatGPT for Clinicians, and other services for customers who are businesses and developers”.^[1] Individuals use ChatGPT under the separate Terms of Use.^[13] Two metric families apply: per-user seats (with optional credits) for ChatGPT plans, and per-token consumption for the OpenAI API. OpenAI states that “OpenAI APIs are billed separately from ChatGPT Plus, Business, Enterprise and Edu.”^[2] The article describes the position as published on 2026-09-26. It follows the general model in Subscription and consumption licensing.

Plans

OpenAI’s pricing FAQ says the paid ChatGPT plans (Go, Plus, Business and Enterprise) “are priced per user per month”. Go, Plus and Business can be billed monthly, and Business and Enterprise annually.^[2]

Business seats are Standard or Premium, and a workspace can mix them. Premium “includes 5x more usage than Standard seats, no 5-hour usage limit”.^[4] Credit-based Enterprise agreements add a Codex seat that provides “access to Codex only”. Under those agreements it is usage-based, “with no fixed cost per user per month”.^[7]

Metrics

The catalog records nine OpenAI metrics under OpenAI’s own names. It gives no conversion between seats, credits and tokens because OpenAI publishes none. The Help Center notes that a previous credit total “is not a direct estimate of the same activity’s dollar cost.”^[6]

For scale, a few flagship list prices at Standard processing, as of 2026-09-26: GPT-6 Astra costs $10.00 input, $1.00 cached input and $50.00 output per 1M tokens. GPT-6 Sol costs $2.00, $0.20 and $10.00. GPT-6 Luna costs $0.10, $0.01 and $0.50. These are “standard processing rates for context lengths under 272K”.^[2] The full table, with long-context and cache-write columns, is on the API pricing page.^[3] The catalog does not copy it.

Counting and metering

ChatGPT Business seats

Each Business workspace has a “2 seat minimum” (rule ChatGPT Business 2 seat minimum, dated 2026-09-26).^[4] The billed quantity is the number of purchased seats, not the number of people using them: “At the start of each billing cycle, you are charged for all standard and premium seats in your workspace, whether or not they’ve been assigned to individual members” (ChatGPT Business bills all purchased seats each cycle).^[4]

• Adding seats triggers an immediate prorated charge. On annual plans, added seats stay committed until the end of the annual term.
• Reductions take effect at the start of the next billing cycle. A workspace cannot reduce below its number of occupied seats.
• An unused paid seat can be assigned at no extra charge. Removing a member frees the seat but does not reduce the paid seat count (ChatGPT Business seat reassignment).^[4]

The Services Agreement requires that End User Accounts “may only be provisioned to, registered for, and used by, a single End User”. It also forbids sharing login credentials between users (One End User per End User Account, effective 2026-01-01).^[1] The Named user licensing article covers this general pattern.

ChatGPT Enterprise and Edu

On credit-based agreements, Enterprise and Edu workspaces “purchase a shared credit pool at the contract level”. Qualifying added seats “are generally reconciled through quarterly true-ups”, counted from the service-term start date rather than calendar quarters (ChatGPT Enterprise credit-based quarterly true-up).^[5]

On token-based agreements, usage across Chat, Work and Codex is metered in tokens and charged in USD at contracted rates. Usage charges are separate from any seat fees. “The number of assigned seats at the start of each billing month sets your baseline seat charge”. Usage and seat fees are invoiced monthly in arrears, and API usage is invoiced separately (ChatGPT Enterprise token-based seat baseline).^[6] Token-based billing “is not currently offered for ChatGPT Edu or ChatGPT for Healthcare.”^[6]

OpenAI API

The Services Agreement says usage-based Fees are charged “based on the usage calculated by OpenAI”. Price changes on the Pricing Page take effect “fourteen days after they are posted” (Usage-based fees calculated by OpenAI).^[1] Playground usage is billed “at the per-token input and output prices”.^[2]

• Batch API: saves 50% on inputs and outputs for tasks run asynchronously over 24 hours (Batch API 50% discount).^[2]
• Data residency: regional processing endpoints carry “a 10% uplift for models released on or after March 5, 2026” (API data residency 10% uplift).^[3]
• Rate limits: measured in RPM, RPD, TPM, TPD and IPM, and set “at the organization level and at the project level, not user level”.^[10]
• Usage tiers: OpenAI moves customers up automatically as spend grows, from Tier 1 ($5 paid, $100 per month usage limit) to Tier 5 ($1,000 paid, $200,000 per month) (API usage tiers graduate automatically with spend).^[10]
• Monthly budget: a customer can set a budget in billing settings, but OpenAI warns that “There may be a delay in enforcing the limit, and you are responsible for any overage incurred.”^[2]

Scale Tier sells “token units” of input and output tokens per minute for one model snapshot. “Each token unit is purchased for a minimum of 30 days.” Billable tokens per minute are averaged over 15-minute intervals, and usage above the entitlement is billed at pay-as-you-go Fast mode rates (Scale Tier 30-day minimum and PAYG overage).^[12] Scale Tier covers models released before GPT-5.6. For later models, OpenAI points to Reserved Tier.^[12]

Data use and retention

Under Services Agreement §4.2, “OpenAI will not use Customer Content to develop or improve the Services, unless Customer explicitly agrees to such use.” Under §4.1, the customer retains ownership of Input and owns Output (No use of Customer Content to improve Services).^[1] The Enterprise privacy page applies the no-training-by-default commitment to ChatGPT Business, Enterprise, for Healthcare, Edu, for Teachers and the API Platform.^[8]

• API: OpenAI states that “As of March 1, 2023, data sent to the OpenAI API is not used to train or improve OpenAI models”. Abuse monitoring logs are “retained for up to 30 days”. Eligible customers can use Zero Data Retention or Modified Abuse Monitoring (API abuse-monitoring retention up to 30 days).^[11]
• ChatGPT Enterprise, Edu and for Healthcare: workspace admins control retention. Deleted conversations are removed within 30 days unless OpenAI is legally required to keep them (ChatGPT business-plan retention and deletion, dated 2026-01-08).^[8]
• ChatGPT Business: deleted or unsaved conversations are removed within 30 days, subject to legal and safety exceptions.^[8]
• Individual plans: OpenAI “may use Content to provide, maintain, develop, and improve our Services”. Users can opt out of model training (Consumer content use and training opt-out).^[13]

Usage restrictions

Services Agreement §3.3 bars using Output “to develop artificial intelligence models that compete with OpenAI’s products and services” except for a Permitted Exception. It also bars trading API keys with third parties and circumventing rate limits or Usage Limits (Output and rate-limit restrictions).^[1] The Usage policies (effective 2025-10-29) apply across OpenAI products. Breaking them “may mean you lose access to our systems” (Usage Policies apply to all OpenAI services).^[9] The Pro help article restates Terms of Use prohibitions on sharing credentials and “Reselling access or using ChatGPT to power third-party services.”^[15] For general compliance method, see License compliance.

Cloud and marketplace availability

The OpenAI documents retrieved for this article do not describe listings on third-party cloud marketplaces. The Services Agreement allows Fees to be paid to “OpenAI or Customer’s reseller”.^[1] For regional processing, ChatGPT Enterprise lists data residency in ten regions.^[2] The API documentation lists the United States, Europe, Australia, Canada, Japan, India, Singapore, South Korea, the United Kingdom and the United Arab Emirates.^[11]

Programs

Out of scope

This article does not cover:

• ChatGPT for Healthcare, ChatGPT for Clinicians and ChatGPT for Financial Services beyond the billing notes above
• ChatGPT Go, which appears only in the pricing FAQ
• ChatGPT Ads
• Service Credit Terms, the Data Processing Addendum and the Service-Specific Terms, which were not retrieved
• Reserved Tier pricing and Fast mode rates
• Image, voice and tool prices
• OpenAI models offered through third-party clouds, which the OpenAI documents retrieved do not describe

Consumer terms for the EEA, Switzerland and UK are a separate Europe Terms of Use. This article cites the version for other regions.^[13] For the general nature of a software license, see that article.

Catalog rows

Sample catalog rows for this article, grouped by table. All rows are on the OpenAI vendor page.

• Metrics
  • Standard seat
  • Premium seat
  • Credits
  • Input tokens
  • Output tokens
  • Token unit
• Rules
  • ChatGPT Business 2 seat minimum
  • ChatGPT Business bills all purchased seats each cycle
  • ChatGPT Enterprise token-based seat baseline
  • No use of Customer Content to improve Services
  • API abuse-monitoring retention up to 30 days
  • API usage tiers graduate automatically with spend
  • Scale Tier 30-day minimum and PAYG overage
• Programs
  • ChatGPT Business
  • ChatGPT Enterprise
  • Scale Tier