Actual OpenAI Pricing 2026

How ChatGPT pricing works

OpenAI runs two separate billing paths: subscriptions for ChatGPT access and usage-based billing through the API. Most business buyers evaluate both because the right approach depends on whether employees are using ChatGPT directly or the organization is building AI into applications, agents, and automated workflows. API usage is billed separately from ChatGPT subscriptions.

Subscription billing

ChatGPT subscriptions charge a recurring fee per user, with model access, usage limits, and features varying by plan. OpenAI offers plans for individuals as well as ChatGPT Business and Enterprise for organizations. Business pricing varies by seat type and billing commitment, while Enterprise pricing is customized based on an organization's requirements and usage. See OpenAI's current pricing page for the latest plan prices and features.

API billing

OpenAI's API uses usage-based pricing, with costs varying by model and workload. Pricing can include input tokens, cached input, cache writes, and output tokens, with different rates depending on the model and context length. OpenAI also offers different processing options that can trade off cost, speed, and availability.

API pricing changes frequently as OpenAI introduces new models. See OpenAI's API pricing page for current per-token rates.

Key cost drivers

Several factors determine your actual OpenAI bill beyond base pricing:

• Model selection. API prices vary significantly across OpenAI models. Using the least expensive model that reliably meets your quality and performance requirements can materially reduce API costs.
• Context accumulation. In multi-turn conversations, each request resends prior turns as context. A long chat or agent loop can cost many times more than the first message alone.
• Reasoning tokens. Reasoning-mode models generate internal reasoning tokens that bill at output rates, even though they aren't shown in the response. Heavier reasoning means a higher effective output bill.
• Prompt caching. Reusing common prompt content can reduce the cost of repeated inputs. OpenAI prices cached input below standard input rates for supported models, although cache writes may carry a separate charge depending on the model.
• Batch processing. OpenAI offers different processing options for API workloads. Batch processing can reduce token costs for asynchronous workloads, while other service tiers prioritize speed or flexibility.

Annual vs. monthly billing

Billing options vary by ChatGPT plan. ChatGPT Business offers lower per-user pricing with an annual commitment than with monthly billing, while Enterprise pricing is customized through OpenAI's sales team. Organizations should compare the subscription price with expected usage and any additional usage-based costs when evaluating total spend.

See OpenAI's pricing page for current monthly and annual rates.

Free tier and trials

OpenAI offers a free version of ChatGPT with lower usage limits and more limited access than its paid plans. API usage is billed separately from ChatGPT subscriptions. OpenAI does not prominently advertise a standard free trial for its paid business plans, so organizations should check OpenAI's pricing page for current offers and eligibility.

‍

ChatGPT pricing plans

OpenAI offers several ChatGPT subscription plans for individuals and organizations, ranging from a free plan to paid individual subscriptions, ChatGPT Business, and custom-priced ChatGPT Enterprise. Paid ChatGPT plans are generally priced per user per month, while OpenAI API usage is billed separately based on usage.

The table below summarizes OpenAI's current ChatGPT plans and starting prices. Features, usage limits, and model access change frequently, so check OpenAI's pricing page for the latest plan details.

Prices shown are based on OpenAI's published pricing as of August 2026 and may change. OpenAI frequently updates plan features, usage limits, and model availability. Visit OpenAI's pricing page for the latest details.

ChatGPT Business pricing

ChatGPT Business starts at $20 per user per month for Standard seats when billed annually, or $25 per user with monthly billing. Premium seats cost $100 per user per month annually or $125 monthly and provide 5x more usage than Standard seats with no five-hour usage limit. Organizations can mix Standard and Premium seats within the same workspace.

ChatGPT Business is designed for organizations that need a secure, managed workspace for employees. It includes centralized billing and administration, usage analytics and spend controls, SAML SSO, MFA, connections to workplace tools, and access to OpenAI's current models. OpenAI does not train its models on Business customer data by default.

See OpenAI's Business pricing page for current features, model access, and usage limits.

ChatGPT Enterprise pricing

ChatGPT Enterprise uses custom pricing and does not have a publicly listed per-user price. Organizations must contact OpenAI's sales team for a quote. OpenAI offers both credit-based and token-based pricing for Enterprise plans, along with volume discounts.

Enterprise is designed for organizations that need additional security, administration, privacy, and support beyond ChatGPT Business. Features include SCIM provisioning, Enterprise Key Management, role-based access controls, expanded context windows, custom data retention policies, data residency, and 24/7 priority support. OpenAI also offers SLAs, custom legal terms, and access to AI advisors for eligible customers.

Because Enterprise pricing is negotiated, the actual cost can depend on deployment size, usage, contract structure, and negotiated terms. Organizations should evaluate these costs alongside any separate OpenAI API spend when estimating their total OpenAI cost.

OpenAI API pricing

OpenAI API pricing is usage-based rather than a fixed monthly subscription. Costs vary by model and can include input tokens, cached input, cache writes, and output tokens. OpenAI's current flagship GPT-5.6 models are Sol, Terra, and Luna.

Current prices per million tokens are:

Short-context pricing

Long-context pricing

GPT-5.6 Sol pricing is promotional and is available through at least November 21, 2026, according to OpenAI. API models and prices change frequently. See OpenAI's API pricing page for the complete model catalog and current rates.

Long-context workloads cost more. For example, GPT-5.6 Sol increases to $8 per million input tokens and $30 per million output tokens for long-context processing. OpenAI also charges separately for capabilities and tools such as realtime audio, image generation, web search, file search, and code execution.

Actual API costs depend on more than token volume. Model selection, context length, caching, reasoning, tool usage, and processing method can all affect total spend. OpenAI also offers different processing options that trade off cost, latency, and availability.

‍

How to negotiate lower OpenAI pricing

OpenAI's individual ChatGPT plans have published prices, but organizations purchasing ChatGPT Business, Enterprise, or significant API capacity may have more opportunities to optimize their total cost. The strongest negotiation strategy depends on your deployment size, expected usage, contract structure, and willingness to commit spend.

1. Commit to annual billing for an immediate discount

OpenAI discounts both Standard and Premium ChatGPT Business seats for annual commitments. Standard seats currently cost $20 per user per month annually versus $25 monthly, while Premium seats cost $100 per user per month annually versus $125 monthly.

If your organization expects to use ChatGPT Business for a full year, compare the savings from an annual commitment against the flexibility of monthly billing before choosing a contract structure.

2. Negotiate Enterprise volume discounts

OpenAI offers volume discounts for ChatGPT Enterprise. If you're deploying ChatGPT across a large workforce, use your expected seat count and adoption plan as leverage when negotiating pricing.

Ask how pricing changes at different deployment levels and whether additional commitments can lower your effective per-user or usage cost. Benchmark the quote against your expected adoption so you aren't committing to more capacity than employees are likely to use.

3. Benchmark OpenAI against alternatives

Before negotiating, compare OpenAI's proposal with both market pricing and competing AI providers. Quotes from alternatives such as Anthropic Claude, Google Gemini, or Microsoft Copilot can provide useful context when evaluating whether OpenAI's pricing is competitive.

Don't compare subscription prices alone. Consider expected usage, model access, API commitments, contract flexibility, and any additional costs that affect the total value of the agreement.

4. Time your purchase around quarter-end

Like most SaaS vendors, OpenAI's sales team works on quarterly targets. Engaging in negotiations toward the end of a fiscal quarter (typically March, June, September, December) can increase your chances of receiving concessions, whether that's a reduced per-seat rate, additional usage credits, or waived onboarding fees.

5. Negotiate Enterprise contract terms beyond price

If OpenAI's headline price is difficult to move, look for other contract terms that can improve the overall value and flexibility of the agreement. Define your requirements upfront so you're negotiating around the security, support, usage, and deployment capabilities your organization actually needs.

Review terms such as payment schedules, renewal provisions, price increases, usage commitments, data retention, SLAs, and pricing for additional seats or capacity. If you expect your OpenAI usage to grow, negotiate how pricing will change as your deployment expands rather than waiting until you need additional capacity.

Organizations using both ChatGPT and the OpenAI API should also consider negotiating their total OpenAI spend together. Consolidating seat and API commitments may create additional leverage when discussing pricing and contract terms.

Finally, weigh any additional discount against the flexibility you're giving up. A larger or multi-year commitment may strengthen your negotiating position, but it can also create unnecessary spend if your AI usage, model preferences, or provider strategy changes during the contract term.