OpenAI API pricing in 2026: GPT-6 token costs and what a monthly bill looks like

OpenAI API plans and prices

The OpenAI API has no seats and no plans to compare. You pay for every token in and out, so the bill depends on which model engineers pick, how long prompts and answers run, and which processing mode and tools each call uses. That makes it easy to start and hard to forecast.

This guide prices the current GPT-6 lineup and GPT-5.5 from OpenAI's official developer pricing page. ChatGPT subscriptions are a separate product with their own guide.

Prices below were recorded on 12 October 2026. The headline figure is the lowest price you can get for that plan. Check the official page before you buy, because vendors change prices without notice.

Read this first. openai.com/api/pricing returned an error, so prices were read on OpenAI's developer pricing page on 12 October 2026. Third-party sources disagree on the GPT-6 Sol and GPT-5.6 Sol naming and on whether one rate is promotional, so this page uses the official table only. Long-context surcharges are confirmed for GPT-5.5 and GPT-5.4 (above 272K input tokens), and not confirmed for the GPT-6 family. The calculator uses input-token prices only, while the scenarios include output. The Zylo benchmark sample size is not stated. Model names change often, so recheck them before you buy.

OpenAI API cost calculator

Move the slider to see what each plan costs at your size. The figures use the list prices on this page and leave out tax, add-ons and any discount you negotiate.

What one app costs at 100M input and 20M output tokens a month

The API has no seats, so this table follows one app at 100M input and 20M output tokens a month through three model and mode choices.

Where the bill grows

A token bill is shaped by choices engineers make in code. These are the ones that move it.

What each OpenAI API plan includes

Compare the plans side by side. Where a cell says Check page, the vendor's pricing page was not specific enough for us to fill it in.

Which OpenAI API plan should you pick?

How to pay less for OpenAI API

1. Route by task

   Send simple calls to GPT-6 Luna and keep Astra for the few tasks that need it.

2. Use Batch and Flex

   Non-urgent jobs cost 50% less on Batch or Flex.

3. Design for caching

   Cached input on GPT-6.1 Sol costs USD 0.10 per 1M tokens against USD 2 uncached.

4. Keep Fast mode targeted

   Enable Fast mode only on latency-critical endpoints, since it doubles GPT-6 rates.

5. Cap output length

   Output tokens cost 5x input, so set maximum output lengths and trim verbose responses.

6. Set project budgets

   Give each team its own project and budget so spikes are caught in days and not at month end.

Subject: OpenAI API renewal on [renewal date]

Hi [account manager],

Our OpenAI API [plan] subscription renews on [renewal date]. Over the last 90 days, [active million input tokens per month] of the [paid million input tokens per month] we pay for used it.

We would like to renew with [target million input tokens per month] on a [term] term. Before we sign, please send:
1. The current list price and the price you are offering us, line by line.
2. The cap on any price increase at the next renewal.
3. The notice period if we want to cut million input tokens per month or move to another plan.

We are comparing this against [alternative] and plan to decide by [decision date]. If you can offer [target discount] percent off list, or a lower price for a longer term, we are ready to talk this week.

Thank you,
[name]
[company]

OpenAI API alternatives and what they cost

If this vendor does not fit, these are the closest options and what they charge.

Keeping OpenAI API spend under control

Token billing makes every engineer a purchaser. A prompt that grows, a retry loop or a new default model changes the invoice, and the person who wrote it rarely sees the number.

• Set budgets and alerts. Usage bills grow without anyone signing anything. Set a monthly budget and an alert at 80 percent of it.
• Tag cost to a team. Give every project an owner so a bill can be traced. Our spend management software guide covers the controls to look for.
• Review commitments before renewal. Committed spend discounts only pay off if you use the commitment. Compare 12 months of real use with the plan before you renew, using a renewal tracker.
• Count the full cost, not the list price. Support, egress, extra seats and add-ons change the real figure. Our total cost of ownership guide shows how to add it up.

Bottom line. Default to GPT-6.1 Sol, route simple work to Luna, keep Astra and Fast mode for the tasks that earn them, and push everything that can wait through Batch. Then give each project a budget and an owner.

OpenAI API pricing questions

More pricing guides

Browse all pricing guides