ChatGPT for Business: Plans, Pricing & How to Manage It

ChatGPT Plans in 2026: What Finance Teams Need to Know

If you manage software budgets at a growing company, you have probably already noticed ChatGPT charges appearing on expense reports, corporate cards, and reimbursement requests. You are not alone. ChatGPT is now the single most widely adopted AI tool in business, and its pricing structure has evolved rapidly since launch. OpenAI now offers six distinct tiers, ranging from free to custom-priced enterprise contracts, and choosing the wrong one (or failing to choose at all) can cost your company thousands per year in duplicate spend and unmanaged subscriptions.

This guide breaks down every ChatGPT plan available in 2026, compares the options that matter for companies with 30 to 500 employees, exposes the hidden costs most organisations miss, and shows you how to get ChatGPT spend under control.

One important note before we start: OpenAI renamed the ChatGPT Team plan to ChatGPT Business in August 2025. If you see references to "Team" elsewhere online, it is the same plan with the same pricing. The name changed; the product did not.

Free: $0/month

The free tier gives individual users access to GPT-4o with limited message allowances. It includes basic features like web browsing, file uploads, and image generation, but with strict usage caps. There is no admin console, no data privacy controls, and no way to manage it centrally. For a company, the free tier is fine for occasional personal use, but it offers zero governance or visibility.

Go: $8/month per user

Launched in 2025 and now available in over 170 countries, Go sits between Free and Plus. It provides expanded access to GPT models with higher message limits than the free tier, more file uploads, and more image generation. However, Go does not include advanced reasoning ("thinking") models, deep research capabilities, or custom GPTs. OpenAI has indicated it may run ads in the Go interface, which makes it a poor fit for professional use. This plan is designed for light individual users who want more than the free tier but do not need the full Plus feature set. For companies, Go is not a viable option because it lacks any admin or governance features.

Plus: $20/month per user

Plus is the standard paid individual plan. It includes access to GPT-4o, reasoning models, DALL-E image generation, Advanced Voice, deep research, and custom GPTs. Message limits are generous but not unlimited. For most knowledge workers, Plus is the default upgrade from Free. The problem for companies: Plus is an individual subscription with no admin controls, no centralised billing, and no data privacy guarantees beyond the standard consumer terms.

Pro: $200/month per user

Pro is built for power users, researchers, and engineers who consistently hit Plus limits. It includes unlimited access to all models, a doubled context window (approximately 680 pages), 250 deep research runs per month, and priority access to new features like Codex and Operator. At $200 per seat, Pro makes sense for a small number of heavy users. The risk for finance teams: employees upgrading themselves to Pro without approval creates $2,400 per year in unbudgeted spend per person.

Business (formerly Team): $25/month per user (annual) or $30/month (monthly)

Business is OpenAI's self-serve plan for organisations. It requires a minimum of two seats and includes a shared workspace, admin console, unified billing, SAML SSO, and MFA. Business data is excluded from model training by default. Users get generous access to frontier models including GPT-4o and reasoning capabilities, plus the ability to create and share custom GPTs within the workspace. Business also supports integrations with tools like Slack, Google Drive, SharePoint, and GitHub. For most companies between 30 and 500 employees, Business is the right starting point.

Enterprise: custom pricing (estimated $45 to $75/month per user)

Enterprise is a sales-negotiated contract with a minimum of around 150 seats. At the lower end, that puts the annual floor at approximately $108,000. Enterprise adds SCIM provisioning, encryption key management (EKM), domain verification, user analytics, role-based access controls, and custom data retention policies. Enterprise pricing depends on seat count, contract length, and negotiated terms. For companies under 500 employees, Enterprise is typically overkill unless you have specific compliance requirements that mandate features like EKM or custom data retention.

Business vs Enterprise: Which Plan Do You Need?

For companies with 30 to 500 employees, the decision usually comes down to Business or Enterprise. Here is how the two plans compare across the features that matter most to finance and IT teams.

The right plan for 30 to 500 employee companies

Most companies in this range should start with Business. It provides the essentials: centralised billing, SSO, data exclusion from training, and a shared workspace. The $25 per user per month price point (annual billing) is predictable and manageable.

Consider Enterprise when you need specific compliance features (EKM, custom data retention, SCIM provisioning) or when your user count approaches 150 or more and you can negotiate volume discounts that bring per-seat costs closer to the Business rate. Some companies also move to Enterprise when they need dedicated account management or custom integrations.

When to upgrade from Business to Enterprise

Three signals that it is time to talk to OpenAI's sales team: your security team requires encryption key management, you need SCIM provisioning for automated user lifecycle management, or your seat count has grown beyond 150 and you want to negotiate better per-seat pricing. If none of these apply, Business is likely sufficient.

The Hidden Costs of ChatGPT at Scale

The sticker price of $25 per user per month looks straightforward. In practice, companies often spend far more than they expect on ChatGPT. Here is where the hidden costs accumulate.

Duplicate subscriptions

This is the most common source of waste. Employees sign up for personal Plus accounts at $20 per month, expensing them or paying out of pocket. When the company later rolls out a Business plan, those individual subscriptions do not automatically cancel. You end up paying $25 per month for the Business seat and $20 per month for the personal Plus account, totalling $45 per user per month for what should cost $25. According to Cledara's research on shadow IT, unauthorised software makes up a significant portion of total software usage during working hours, and ChatGPT is now the single most common shadow IT application in the workplace.

Uncontrolled Pro upgrades

The Pro plan at $200 per month is 8x the cost of Business. When an engineer or researcher upgrades to Pro without going through procurement, it creates $2,400 per year in unplanned spend. Multiply that across even five self-upgrading power users and you are looking at $12,000 annually that never appeared in a budget. This problem is especially common in engineering and data science teams, where individuals feel justified upgrading because they hit message limits frequently. Without a clear policy and enforcement mechanism, Pro creep is inevitable.

API usage costs on top of seat licences

ChatGPT seat licences and OpenAI API usage are billed separately. Development teams using the API for internal tools, automations, or product features generate consumption-based charges that can spike unpredictably. A single poorly optimised workflow calling GPT-4o can generate hundreds of dollars in API charges per month, on top of existing seat costs. The danger here is that API costs are invisible to most finance teams until the invoice arrives. Unlike seat-based pricing, which is predictable, API consumption scales with usage and can double or triple month-over-month as teams build more AI-powered workflows.

Custom GPTs and plugin proliferation

Business and Enterprise plans allow teams to build custom GPTs. While individually these are low-cost, they encourage broader adoption and higher usage. More importantly, custom GPTs often connect to third-party data sources and plugins, each of which may carry its own subscription cost. The ChatGPT line item stays flat, but the ecosystem of connected tools around it grows. Finance teams should treat ChatGPT not as a single line item but as a platform that generates downstream software spend.

How to Manage ChatGPT Spend Across Your Company

Getting ChatGPT costs under control requires the same discipline you would apply to any other software category. Here are the practical steps.

Consolidate onto one plan

Start with a full audit. Pull expense reports and corporate card statements for the past six months and search for any charges from OpenAI. Look for "ChatGPT," "OpenAI," and "GPT" across all payment methods, including personal reimbursements. Every individual Plus or Pro subscription you find should be migrated to the company Business plan and the personal subscription cancelled. OpenAI's help documentation covers how to cancel individual subscriptions after migrating to a company plan. Do not underestimate the scale of this problem: at many companies, the number of personal ChatGPT subscriptions exceeds the number of managed seats by a significant margin.

Set clear purchasing policies

Establish a policy that all AI tool subscriptions, including ChatGPT, must go through a centralised approval process. This prevents employees from signing up for individual plans or upgrading to Pro without authorisation. Define which roles genuinely need Pro-tier access ($200 per month) versus standard Business access ($25 per month), and build that into your approval workflow. Communicate the policy clearly: most employees who sign up for personal ChatGPT accounts are not trying to circumvent policy. They simply do not know a company plan exists or how to access it. A brief internal announcement with clear instructions can eliminate most shadow subscriptions.

Track API usage separately

If your engineering team uses the OpenAI API, connect the API account to a monitoring tool that tracks daily spend by project or team. Set budget alerts at 80% of your monthly target so you catch overruns before they hit the invoice. API costs are consumption-based and can fluctuate significantly month to month, especially as teams experiment with new models or scale up existing workflows. Treating API spend as a separate budget line (not lumped in with ChatGPT seat costs) gives you the granularity to spot issues early.

Right-size licences quarterly

Review usage data every quarter. Identify seats that have been inactive for 30 or more days and reassign or remove them. On a Business plan at $25 per user per month, every unused seat costs $300 per year. For a 100-person deployment, even 15% inactive seats represent $4,500 in annual waste. Do not assume that every employee who requested a seat six months ago still needs one. Roles change, projects end, and tools fall out of favour. Quarterly licence reviews are standard practice for established SaaS categories like CRM and project management. Apply the same rigour to AI tools.

Separate ChatGPT from other OpenAI charges

OpenAI bills ChatGPT subscriptions and API usage through different channels, but they can appear on the same invoice or card statement. Make sure your finance team can distinguish between the two. Use separate payment methods or virtual cards for each to maintain clear cost attribution. This separation is critical for accurate budgeting and for understanding your true cost per user. Without it, a $3,000 monthly ChatGPT bill might actually be $1,500 in seats and $1,500 in API usage, each requiring a different management approach.

How Cledara Manages ChatGPT (and All Your AI Tools)

ChatGPT is one of many AI tools your company is likely paying for. Cledara's data shows that AI spending has grown dramatically, with ChatGPT seeing 33x more usage than the next closest AI tool. Meanwhile, emerging players like Perplexity and Claude are growing at rates above 200%, meaning your AI tool portfolio is expanding rapidly whether you planned for it or not. Managing ChatGPT in isolation misses the bigger picture. Here is how Cledara helps you control the full AI spend landscape.

AI Dashboard: real-time visibility into usage-based AI spend

Cledara's AI Dashboard connects directly to your OpenAI developer API key and tracks daily spend with visual breakdowns by team, project, or environment. You can set budget thresholds and receive alerts when spend approaches your limit. This solves the API cost visibility problem that catches many finance teams off guard. Note: the AI Dashboard tracks usage-based API spend; for ChatGPT seat subscriptions, Cledara's virtual card controls (below) provide the governance layer.

Virtual cards with automatic spend limits

Cledara issues a dedicated virtual card for your ChatGPT Business subscription with a specific spend limit. If OpenAI tries to charge more than the approved budget (for example, if seats are added without approval), the transaction is declined. Smart top-ups adjust the card balance to match actual billing amounts, so you are not tying up cash unnecessarily while still maintaining hard spending caps.

Engage: discover employees with personal ChatGPT subscriptions

Cledara's Engage browser extension detects which SaaS tools employees are actually using, including ChatGPT accessed through personal accounts. This is how you find the shadow AI subscriptions that do not appear on any corporate card. Once identified, you can migrate those users to the company Business plan and eliminate duplicate spend. Engage tracks SaaS provider URLs only, not browsing history, so employee privacy is maintained.

Benchmarks: know if you are overpaying

Cledara's Benchmarks feature compares your ChatGPT per-seat cost against anonymised data from companies of similar size. If you are paying above the 75th percentile, that is a signal to renegotiate or reassess your plan tier. Combined with utilisation data from Engage, benchmarks give you the evidence you need to right-size your ChatGPT investment.

One-click cancellation when it is time to switch

If your company decides to move from ChatGPT Business to Enterprise (or switch AI providers entirely), Cledara makes the transition clean. Because Cledara controls the payment card, cancelling is as simple as freezing it. Payments stop instantly. No vendor runaround, no forgotten renewals, no charges continuing months after you thought you cancelled. This is especially valuable in the fast-moving AI landscape, where your team might decide to switch from ChatGPT to Claude, Gemini, or a specialised vertical AI tool. With Cledara, you can spin up a new virtual card for the replacement tool and freeze the old one in minutes.

Negotiation Copilot: data-driven renewal leverage

When your ChatGPT contract comes up for renewal, Cledara's Negotiation Copilot arms you with the data you need. It pulls your actual spend history, benchmark comparisons showing what similar companies pay, and utilisation rates from Engage. It can even generate a pre-filled negotiation email with data-backed arguments for a better rate. Whether you are renewing a Business plan or negotiating an Enterprise contract, having real pricing data from across the market gives you leverage that a spreadsheet cannot.

ChatGPT is a powerful tool, but like any software that scales across an organisation, it requires active management. The companies that get the most value from AI are not the ones spending the most; they are the ones who know exactly what they are spending, on which plans, for which users, and whether the investment is delivering results. That is the visibility Cledara provides.