By ChatGPT Price Sheet: Plans, API and Business
Every ChatGPT price on one page: Free to Pro plans, Business seats and API token rates, read from OpenAI's own pages, plus how billing and tax work.
ChatGPT prices start at $0 for Free, $8 per month for Go and $20 per month for Plus. Pro costs $100, $200 or $500 per month. Business seats start at $20 per user per month on annual billing, and the API (application programming interface) charges per token, a small piece of text that a model reads or writes. Every figure here is in US dollars and comes from OpenAI's own pages, read on 5 November 2026.
ChatGPT plan prices
The data behind this chart
[
{
"plan": "Free",
"usd_per_month": 0,
"notes": "No payment."
},
{
"plan": "Go",
"usd_per_month": 8,
"notes": "Monthly billing only. Ads are being tested with eligible Go users."
},
{
"plan": "Plus",
"usd_per_month": 20,
"notes": "Monthly billing only."
},
{
"plan": "Pro 100",
"usd_per_month": 100,
"notes": "Monthly billing only. Promotional API credits available."
},
{
"plan": "Pro 200",
"usd_per_month": 200,
"notes": "Monthly billing only. Promotional API credits available."
},
{
"plan": "Pro 500",
"usd_per_month": 500,
"notes": "Monthly billing only. Promotional API credits available. Includes Ultrafast mode."
}
]The ChatGPT pricing page shows three Pro usage tiers. The Pro help article lists the three tiers. Some countries have local-currency prices, so your checkout can show a different amount.
Ultrafast mode is on Pro 500 and on eligible Enterprise and Edu plans. It is not on Pro 100, Pro 200 or Plus.
This page lists prices only. For the limits, read ChatGPT Plus price and limits. To pick a model inside ChatGPT, read GPT-6 Astra vs Sol vs Luna in ChatGPT. To compare vendors, read Claude vs ChatGPT subscription cost.
ChatGPT Business seat prices
The data behind this chart
[
{
"plan": "Standard seat",
"usd_billed_annually": 20,
"usd_billed_monthly": 25
},
{
"plan": "Premium seat",
"usd_billed_annually": 100,
"usd_billed_monthly": 125
}
]The Business pricing page shows these prices. OpenAI's help article gives them "for most countries", so the price in your country may differ.
A Business workspace needs at least 2 Standard or Premium seats in total. Some new workspaces can start with 1 paid seat. Availability is limited, and the minimum shown at checkout is the one that applies to you.
The pricing pages show no price for Enterprise or Edu. You contact OpenAI sales, and for Enterprise your agreement sets the billing period and invoice schedule. OpenAI does publish a token rate card, but it applies only to eligible token-based Enterprise agreements.
ChatGPT API prices per 1M tokens
The data behind this chart
[
{
"label": "gpt-6-astra",
"input_usd": "10.00",
"cached_input_usd": "1.00",
"output_usd": "50.00"
},
{
"label": "gpt-6.1-sol",
"input_usd": "2.00",
"cached_input_usd": "0.10",
"output_usd": "10.00"
},
{
"label": "gpt-6.1-terra",
"input_usd": "1.00",
"cached_input_usd": "0.05",
"output_usd": "5.00"
},
{
"label": "gpt-6-luna",
"input_usd": "0.10",
"cached_input_usd": "0.01",
"output_usd": "0.50"
},
{
"label": "gpt-6-sol",
"input_usd": "2.00",
"cached_input_usd": "0.20",
"output_usd": "10.00"
},
{
"label": "gpt-5.6-sol",
"input_usd": "4.00",
"cached_input_usd": "0.40",
"output_usd": "20.00"
},
{
"label": "chat-latest",
"input_usd": "5.00",
"cached_input_usd": "0.50",
"output_usd": "30.00"
},
{
"label": "gpt-5.3-codex",
"input_usd": "1.75",
"cached_input_usd": "0.175",
"output_usd": "14.00"
}
]Short context means up to 272K input tokens. Cached input is the lower rate for input that the API reuses from its prompt cache.
Do not assume one ratio for every model. Output is 5 times input on every row except chat-latest (6 times) and gpt-5.3-codex (8 times). Cached input is one tenth of input on most rows, but one twentieth on gpt-6.1-sol and gpt-6.1-terra.
The gpt-6-sol model still has a price, but its model page points to GPT-6.1 Sol as the newer model. OpenAI says gpt-5.6-sol promotional pricing is available until 30 November 2026. The page does not say what the price becomes after that date.
The data behind this chart
[
{
"label": "gpt-6-astra, Batch or Flex",
"input_usd": "5.00",
"cached_input_usd": "0.50",
"output_usd": "25.00"
},
{
"label": "gpt-6.1-sol, Batch or Flex",
"input_usd": "1.00",
"cached_input_usd": "0.05",
"output_usd": "5.00"
},
{
"label": "gpt-6.1-terra, Batch or Flex",
"input_usd": "0.50",
"cached_input_usd": "0.025",
"output_usd": "2.50"
},
{
"label": "gpt-6-luna, Batch or Flex",
"input_usd": "0.05",
"cached_input_usd": "0.005",
"output_usd": "0.25"
},
{
"label": "gpt-6-astra, Fast",
"input_usd": "20.00",
"cached_input_usd": "2.00",
"output_usd": "100.00"
},
{
"label": "gpt-6.1-sol, Fast",
"input_usd": "4.00",
"cached_input_usd": "0.20",
"output_usd": "20.00"
},
{
"label": "gpt-6.1-terra, Fast",
"input_usd": "2.00",
"cached_input_usd": "0.10",
"output_usd": "10.00"
},
{
"label": "gpt-6-luna, Fast",
"input_usd": "0.20",
"cached_input_usd": "0.02",
"output_usd": "1.00"
},
{
"label": "gpt-6-astra, Ultrafast",
"input_usd": "60.00",
"cached_input_usd": "6.00",
"output_usd": "300.00"
},
{
"label": "gpt-6.1-sol, Ultrafast",
"input_usd": "12.00",
"cached_input_usd": "0.60",
"output_usd": "60.00"
}
]On these models, Batch and Flex cost half of Standard. Fast costs double. Ultrafast costs 6 times Standard, and only gpt-6-astra and gpt-6.1-sol are listed for it. Fast is the new name for Priority processing, which OpenAI renamed on 30 July 2026.
The data behind this chart
[
{
"label": "gpt-6-astra",
"input_usd": "20.00",
"cached_input_usd": "2.00",
"output_usd": "75.00"
},
{
"label": "gpt-6.1-sol",
"input_usd": "4.00",
"cached_input_usd": "0.20",
"output_usd": "15.00"
},
{
"label": "gpt-6.1-terra",
"input_usd": "2.00",
"cached_input_usd": "0.10",
"output_usd": "7.50"
},
{
"label": "gpt-6-luna",
"input_usd": "0.20",
"cached_input_usd": "0.02",
"output_usd": "0.75"
}
]Long context means more than 272K input tokens. For these models, input and cached input double, and output is 1.5 times the short-context rate. Regional processing (data residency) endpoints add 10% for models released on or after 5 March 2026.
All API figures come from the OpenAI API pricing page. That page also lists older models and prompt-cache write prices, which this sheet leaves out.
What does a per-token price mean for one job?
This table applies the Standard long-context prices to one example: 1M input tokens and 200K output tokens, with no caching. The prompt is above 272K input tokens, so these rates apply to the full request. It is arithmetic on list prices, not a measured workload.
The data behind this chart
[
{
"label": "gpt-6-astra",
"input_cost_usd": "20.00",
"output_cost_usd": "15.00",
"total_usd": "35.00"
},
{
"label": "gpt-6.1-sol",
"input_cost_usd": "4.00",
"output_cost_usd": "3.00",
"total_usd": "7.00"
},
{
"label": "gpt-6.1-terra",
"input_cost_usd": "2.00",
"output_cost_usd": "1.50",
"total_usd": "3.50"
},
{
"label": "gpt-6-luna",
"input_cost_usd": "0.20",
"output_cost_usd": "0.15",
"total_usd": "0.35"
}
]The same job costs $35.00 on gpt-6-astra and $0.35 on gpt-6-luna. Output tokens cost more than input tokens, so a job that writes longer answers costs more than this example.
Monthly or annual billing: which plans offer it?
Business is the only self-serve plan with a choice. Go, Plus and Pro are monthly only. Plus has no annual billing, and Pro has no annual billing and no way to prepay several months. Enterprise is sold on annual plans.
The data behind this chart
[
{
"plan": "Standard seat",
"usd_per_year_billed_annually": 240,
"usd_per_year_billed_monthly": 300
},
{
"plan": "Premium seat",
"usd_per_year_billed_annually": 1200,
"usd_per_year_billed_monthly": 1500
}
]One Standard seat for 12 months costs $240 on annual billing and $300 on monthly billing. Annual billing is less flexible. On annual plans, added seats stay committed to the end of the annual term. A Business owner can change the billing schedule, but the change takes effect at renewal. A new seat starts immediately and is prorated, which means you pay for the part of the billing period that is left.
Self-serve subscriptions bill on the date you subscribed, and you cannot change that schedule. Paid plans renew automatically until you cancel.
Is tax added on top of ChatGPT prices?
Check the total at checkout, because the plan price may not be your final charge. The pricing page does not mention tax. OpenAI's Terms of Use for users outside the EEA (European Economic Area), Switzerland and the UK, including US users say: "You’re responsible for all applicable taxes, and we’ll charge tax when required." Separate European terms cover the EEA, Switzerland and the UK. So tax is added where your location requires it.
A Business buyer in a region with VAT (value added tax) can enter a Tax ID at checkout. This exempts the purchase from VAT, and the same status applies to future invoices.
In the US, eligible organizations can request a sales tax exemption through Support. An approved exemption applies only to future invoices. Certificates issued to OpenAI, L.L.C. are not valid from 1 January 2026.
Why are a ChatGPT subscription and API usage billed separately?
A subscription is a flat fee for using ChatGPT. The API is metered: your code sends and receives tokens, and you pay for each one. OpenAI runs two separate billing systems for these, so a paid ChatGPT plan does not cover API calls.
To start paid API usage, add a payment method in your API account billing settings. API credits have a minimum purchase of $5, and the default amount is $10. Purchased credits expire after 1 year. They are non-refundable unless law, a contract or an approved exception says otherwise.
There is one exception, and it is a promotion. Eligible Pro and Business subscribers can activate promotional API credits.
The data behind this chart
[
{
"plan": "Pro 100",
"usd_credit": 100,
"notes": "Per month."
},
{
"plan": "Pro 200",
"usd_credit": 200,
"notes": "Per month."
},
{
"plan": "Pro 500",
"usd_credit": 500,
"notes": "Per month."
},
{
"plan": "Business Standard seat",
"usd_credit": 20,
"notes": "Per paid seat, pooled for the workspace. Cap: $500 per monthly billing period, or $6,000 per annual billing cycle on annual prepaid workspaces."
},
{
"plan": "Business Premium seat",
"usd_credit": 100,
"notes": "Per paid seat, pooled for the workspace. The same cap covers both seat types."
}
]You do not need an API payment method to use an active grant. These credits are separate from ChatGPT usage, so they do not raise your ChatGPT or Codex limits. Plus and free trial subscriptions do not qualify, and Enterprise and Edu are excluded. OpenAI says the benefit may change or be removed. All other API usage on a Business workspace is still billed separately.
What changed recently, and what is only announced?
On 29 October 2026, Pro 100 and Pro 500 got twice their previous Work and Codex allowance. OpenAI also canceled the Pro 200 reduction from its 29 September email.
OpenAI has announced that, starting in December, the usage allowance on Plus, Pro and Business will be shared across ChatGPT and Codex. On 5 November 2026 this is an announcement, not a change in effect. If you use Codex, read how Codex works with a ChatGPT subscription.
FAQ
How much does ChatGPT cost per month?
As of 5 November 2026, Free is $0, Go is $8 and Plus is $20 per month. Pro is $100, $200 or $500 per month. Tax is added where required, and some countries have local-currency prices.
How much is a ChatGPT Business seat?
A Standard seat is $20 per user per month billed annually, or $25 billed monthly. A Premium seat is $100 billed annually, or $125 billed monthly. OpenAI gives these prices for most countries. A workspace normally needs at least 2 paid seats.
Can I pay for ChatGPT Plus annually?
No. OpenAI's help article on Plus says it does not support annual billing, so Plus is $20 billed monthly. Business is the self-serve plan that offers annual billing.
Does a ChatGPT subscription include API usage?
No. ChatGPT and the API platform have separate billing systems, so API usage is billed separately from your subscription. The one exception is a promotion: eligible Pro and Business subscribers can activate promotional API credits, and OpenAI says this benefit may change or be removed.
What does the ChatGPT API cost per 1M tokens?
At Standard short-context prices on 5 November 2026, input and output per 1M tokens are $10.00 and $50.00 on gpt-6-astra, $2.00 and $10.00 on gpt-6.1-sol, $1.00 and $5.00 on gpt-6.1-terra, and $0.10 and $0.50 on gpt-6-luna.