Actual OpenAI Pricing 2026

How ChatGPT pricing works

OpenAI runs two separate billing paths: flat monthly subscriptions for ChatGPT access and per-token billing through the API. Most buyers evaluate both, because the right path depends on whether your usage is human-paced (chat, Projects, Deep Research, Codex in the app) or machine-driven (agents, automation, backend integrations). API access is billed separately and is not included in any ChatGPT subscription.

Subscription billing

ChatGPT subscriptions charge a fixed monthly fee per user with usage limits that vary by tier. Plans run from a free tier through Go ($8/month), Plus ($20/month), two Pro tiers ($100 and $200/month), Business ($20–$25/seat/month), and custom Enterprise pricing. Subscriptions bundle the ChatGPT web and mobile apps and features like Deep Research, Codex, Projects, tasks, and image generation, with model access and limits scaling by tier.

API billing

API pricing charges per token with no monthly ceiling. You pay separately for input tokens (your prompts and context), cached input tokens (repeated context, billed at a steep discount), and output tokens (the model's response, including any internal reasoning). OpenAI's current flagship API models:

• GPT-5.5 (most capable): $5/MTok input, $30/MTok output
• GPT-5.4 (more affordable): $2.50/MTok input, $15/MTok output
• GPT-5.4 mini (strongest small model): $0.75/MTok input, $4.50/MTok output

MTok = million tokens. One token is roughly 4 characters or 0.75 words in English.

Key cost drivers

Several factors determine your actual OpenAI bill beyond base pricing:

• Model selection. GPT-5.5 output costs roughly 6.7x more per token than GPT-5.4 mini and 2x more than GPT-5.4. Picking the smallest model that clears your quality bar is the single biggest lever on API cost.
• 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.
• Cached input. Repeated context (system prompts, documents) is automatically priced as cached input — about 90% cheaper than standard input (for example, $0.50/MTok vs. $5/MTok on GPT-5.5).
• Batch processing. The Batch API applies a flat 50% discount on both input and output for asynchronous workloads completed within 24 hours.

Annual vs. monthly billing

Annual commitments reduce subscription costs on the plans that offer them:

• Business: $20/seat/month (annual) vs. $25/seat/month (monthly) — 20% savings
• Enterprise: billed annually by default; pricing negotiated through sales

Individual plans (Go, Plus, Pro) are monthly-only — there is no annual discount on the consumer tiers, and both Pro tiers ($100 and $200) bill monthly.

API-specific pricing

For developers and teams building on the API, additional costs apply beyond base token pricing:

• Web search: $10 per 1,000 calls; search content tokens are free
• Containers (code execution): 1 GB for $0.03 / 64 GB for $1.92 per container; from March 31, 2026, billed per 20-minute session per container
• Data residency / regional processing: adds a 10% multiplier on token pricing
• Service tiers: Batch (50% off, asynchronous over 24 hours), Priority (reliable high-speed, pay-as-you-go), and Flex (lower cost, slower, for non-production work)

Free tier and trials

ChatGPT Free gives access to the chat interface across web, iOS, and Android with limited messages, limited model access, and (in the US) ads. There is no prominently advertised free trial for paid plans — the Free tier is the entry point for evaluation. API usage is separate and pay-as-you-go; it is not included in any ChatGPT subscription.

‍

ChatGPT pricing plans

OpenAI organizes pricing into consumer plans (individuals), business and enterprise plans (organizations), and a separate per-token API for developers. Below is a breakdown of each.

ChatGPT Free — $0/month

ChatGPT Free gives everyone access to AI-powered chat at no cost, though with significant limitations on usage and model access. This plan may include ads.

• Models: Limited access to GPT-5.5 Instant 
• Messages: Limited messages and uploads
• Features included: Search, Canvas, Voice, Apps, Projects, Study mode, limited image generation, limited deep research, limited Codex access
• Notable limits: No access to GPT-5.5 Thinking or GPT-5.5 Pro; limited memory; no tasks, developer mode, or record mode; slower response times
• Best for: Casual users who want to try ChatGPT or use it occasionally for simple tasks

ChatGPT Go — $8/month

Go expands on the Free tier with more usage capacity, though it still lacks access to advanced reasoning models. This plan may include ads.

• Models: GPT-5.5 Instant (expanded)
• Messages: More messages and uploads than Free
• Features included: Everything in Free plus expanded voice, voice with video, data analysis, vision, create & share GPTs, longer memory
• Notable limits: No GPT-5.5 Thinking or Pro models; limited deep research; limited Codex; bandwidth-limited response times; may include ads
• Best for: Regular users who need more daily capacity but don't require advanced reasoning

ChatGPT Plus — $20/month

Plus is the core paid plan for individuals who need advanced AI capabilities including reasoning models and expanded features.

• Models: GPT-5.5 Instant (expanded), GPT-5.5 Thinking (expanded), GPT-5 Thinking Mini (expanded)
• Messages: Unlimited (subject to reasonable use policies)
• Features included: Everything in Go plus GPT-5.5 Thinking, expanded deep research and agent mode, tasks, projects, custom GPTs, Codex, developer mode, record mode, image generation with Thinking, interactive tables and charts, apps connecting to internal tools, early access to new features
• Notable limits: No GPT-5.5 Pro model; no company knowledge or admin features
• Best for: Professionals, power users, and knowledge workers who rely on ChatGPT daily for complex work

ChatGPT Pro — $100/month

The lower of two Pro tiers (launched April 2026). Individual plan — monthly billing only, no annual discount.

• Usage: Roughly 5x Plus limits
• Models: Full model suite, including GPT-5.5 Pro reasoning 
• Features: Everything in Plus, plus elevated Codex, Deep Research, and agent-mode limits
• Context window: Expanded vs. Plus 
• Best for: Heavy individual users who hit Plus caps regularly but don't need the maximum tier

ChatGPT Pro — $200/month

The maximum individual tier. Monthly billing only.

• Usage: Roughly 20x Plus limits
• Models: Everything in the $100 tier, including GPT-5.5 Pro
• Context window: Largest available
• Features: Maximum Deep Research and agent-mode usage; highest output limits
• Best for: Users who exhaust Plus daily, run parallel workloads, or feed book-length inputs that need the full context window

ChatGPT Business — $20/user/month (billed annually)

A secure, collaborative workspace for teams. Requires a minimum of 2 users.

• Pricing: $20/user/month billed annually; $25/user/month billed monthly
• Models: GPT-5.5 Instant (unlimited), GPT-5.5 Thinking (flexible, credits-based), GPT-5.5 Pro (flexible, credits-based)
• Messages: Unlimited core chat
• Features included: Everything in Plus, plus 60+ app integrations (Slack, Google Drive, SharePoint, GitHub, Atlassian), company knowledge, workspace GPTs, admin console, analytics dashboard, unified billing, bulk member management, domain verification, SAML SSO, SOC 2 Type 2 compliance
• Two seat types available: standard ChatGPT seats and Codex-only seats. Workspaces can mix both. Codex seats provide AI-powered coding capabilities (code generation, security reviews, multi-agent workflows) on flexible usage-based pricing.
• Privacy: No training on your data
• Best for: Startups and growing businesses that need a secure team workspace with admin controls, app integrations, and optional development-focused Codex access.

ChatGPT Enterprise — Custom pricing

Enterprise provides maximum security, compliance, and support for large organizations. Pricing requires contacting OpenAI sales.

• Models: All models with flexible access; fastest response times; 128K GPT Instant context window
• Features included: Everything in Business, plus expanded context windows, SCIM provisioning, Enterprise Key Management (EKM), granular GPT controls and group permissions, role-based access controls, Compliance API Logs Platform, IP allowlisting, data residency in 10 regions (US, EU, UK, JP, CA, KR, SG, IN, AU, UAE), branded workspace, connector registry, ISO 27001/27017/27018/27701 certification
• Support: 24/7 priority support, SLAs, custom legal terms, dedicated onboarding, ongoing account management, custom security review, access to AI advisors
• Privacy: No training on your data; custom data retention policies
• Best for: Large enterprises with strict security, compliance, and data residency requirements

OpenAI API pricing

For developers and teams building on the API, pricing is per token and varies by model. All prices are per million tokens (MTok). Standard processing rates shown; the Batch API applies a 50% discount on input and output, and data residency / regional processing adds 10%.

Flagship models

Standard rates reflect context lengths under 270K tokens.

Multimodal models

Tools and platform costs

Service tiers

OpenAI offers three service tiers for API access, on top of the default standard rate:

• Batch: asynchronous workloads completed within 24 hours, at a 50% discount on input and output
• Priority: reliable, high-speed performance with pay-as-you-go flexibility
• Flex: lower cost in exchange for slower responses and occasional resource unavailability — best for non-production or lower-priority tasks

Enterprise customers can also access Scale Tier and Reserved Capacity for larger workloads.

‍

ChatGPT hidden costs and fees

ChatGPT's listed prices don't tell the full story. Several cost factors are easy to miss when comparing plans.

Ads on the Go plan

The $8/month Go plan may include advertisements. OpenAI states this plan "may include ads," making it the only paid tier with advertising. If an ad-free experience matters to your team, Plus at $20/month is the lowest ad-free option.

Credits for advanced models on Business and Enterprise

Business and Enterprise plans list GPT-5.5 Thinking and GPT-5.5 Pro access as "Flexible" — meaning these models require purchasing additional credits beyond the base subscription. The base per-seat price covers core chat and GPT-5.5 Instant, but heavy use of reasoning models will increase your total cost.

"Unlimited" subject to abuse guardrails

Multiple plans advertise "unlimited" messages, but OpenAI qualifies this: usage must be "reasonable and comply with our policies." Extremely heavy usage could trigger throttling or restrictions, though OpenAI does not publish specific thresholds.

Context window varies by plan

The amount of text ChatGPT can process in a single conversation varies significantly by plan — Free and lower tiers have a much smaller window, while Enterprise reaches 128K tokens (~250 pages) for GPT Instant. If your use case involves long documents, the cheaper plans may not be sufficient.

‍

How to negotiate lower ChatGPT pricing

List prices for ChatGPT's individual plans are fixed, but Business and Enterprise contracts offer meaningful room to negotiate — especially for annual commitments, larger teams, or organizations that qualify for special programs.

1. Commit to annual billing for an immediate discount

The Business plan drops from $25/user/month to $20/user/month when billed annually — a 20% savings that requires no negotiation. If you're currently paying monthly, switching to annual billing is the simplest cost reduction available. For Enterprise contracts, annual commitment is standard and typically a prerequisite for volume discounts.

2. Negotiate volume discounts on Enterprise

OpenAI explicitly mentions "volume discounts" as part of Enterprise pricing. If you're deploying ChatGPT across hundreds or thousands of seats, use your seat count as leverage. Request tiered pricing that decreases per-seat cost as your team scales. Come prepared with a specific headcount commitment and a multi-year timeline to strengthen your position.

3. Apply for Nonprofit or Education pricing

OpenAI offers up to 75% off Business or Enterprise plans for qualifying nonprofits through the OpenAI for Nonprofits program. Universities can access discounted pricing through ChatGPT Edu. If your organization qualifies, this is the largest discount available — contact OpenAI's sales team directly to apply.

4. Use competitor pricing as leverage

ChatGPT competes directly with tools like Google Gemini, Microsoft Copilot, Anthropic Claude, and Perplexity. If you have quotes from competing AI platforms, share them with OpenAI's sales team during Enterprise negotiations. Demonstrating that you're actively evaluating alternatives gives the sales team a reason to offer more competitive terms.

5. 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.

6. Negotiate Enterprise contract terms beyond price

If the per-seat price is firm, negotiate on other levers that reduce your total cost of ownership. Ask for extended payment terms (net-60 or net-90 instead of net-30), custom data retention policies at no extra charge, dedicated onboarding and training hours, SLA commitments with credits for downtime, or the inclusion of additional seats at the contracted rate if your team grows during the contract period.

Negotiating custom Enterprise contracts

Enterprise pricing is fully custom and requires contacting OpenAI's sales team. To get the best deal:

• Define your requirements upfront: Specify the security features you need (EKM, SCIM, data residency, IP allowlisting) so the quote reflects your actual use case rather than a maximum bundle.
• Consolidate spending: If your organization also uses the OpenAI API, negotiate a combined agreement that covers both ChatGPT Enterprise and API usage for better overall terms.
• Push for multi-year discounts: A 2- or 3-year commitment reduces OpenAI's churn risk and gives you leverage to request a lower annual rate.
• Request a pilot period: Ask for a 30–90 day pilot at a reduced rate or with a limited seat count before committing to a full deployment. This lowers your risk and gives OpenAI confidence in your intent to scale.