Skip to main content

ChatGPT Edu Rate Card

Blue banner

ChatGPT Edu Rate Card

ChatGPT Edu (including Codex) is available to University staff and students with standard usage limits.

Enhanced access is available for individuals requiring higher capacity. This is provided on a cost-recovery basis, with charges calculated on actual usage at OpenAI’s standard rates. You can request access to higher limits in ChatGPT Edu through our Payment for Higher Limits form

Access is structured in tiers. Your tier sets your maximum monthly usage limits and therefore your monthly cost ceiling. It is not a fixed subscription fee.

Actual charges are calculated based on real usage and may be lower than the monthly ceiling shown in the rate table below.

Maximum monthly cost

Maximum annual cost

Free tier

Provided free as standard

Provided free as standard

Tier 2

$104 (+VAT)

$1248 (+VAT)

Tier 3

$208 (+VAT)

$2496 (+VAT)

Tier 4

$520 (+VAT)

$6240 (+VAT)


The tables below shows the maximum monthly usage per limited feature, per user within your selected tier. While consumption is measured differently between ChatGPT Edu and Codex, the usage pool is shared between the two apps.

Allowances are not cumulative. Your total usage is limited to the stated monthly cap for each individual feature. These limits do not combine or stack across models and features. These can be allocated flexibly across features.

As best practice, we recommend using GPT-5.6 Luna on Light effort with Standard Speed enabled in Codex and ChatGPT Work, and GPT-5.5 Instant (which does not consume credits) in ChatGPT where appropriate. You may need to click 'Advanced' on the model selector to see all available options. This should allow for the most efficient use of your monthly limits.

ChatGPT Edu

Maximum monthly usage
per limited feature

Free tier

Tier 2

Tier 3

Tier 4

Models (in ChatGPT Edu)

GPT-5.6 Sol / GPT 5.5 Instant

Unlimited

Unlimited

Unlimited

Unlimited

GPT-5.6 Sol Medium / High

50

310

570

1,350

or

or

or

or

GPT-5.6 Sol Pro

No use

62

114

270

Features (in ChatGPT Edu)

Agent

16

103

190

450

or

or

or

or

Deep Research

No use

62

114

270

or

or

or

or

Image Generation

100

620

1,140

2,700

or

or

or

or

Voice (in minutes)

100

620

1,140

2,700

 

ChatGPT Agent Plugins

Maximum monthly token usage for ChatGPT plugin

Plugin

Free tier

Tier 2

Tier 3

Tier 4

ChatGPT for Excel

~9.1M

~56.4M

~103.6M

~245.5M

ChatGPT for PowerPoint

~9.1M

~56.4M

~103.6M

~245.5M

 

Codex and ChatGPT Work

To be granted access to the Codex through the ChatGPT app at Oxford, you must complete the relevant forms:

  1. ChatGPT app Consent form - use this form to request access to the ChatGPT app, giving you availability of ChatGPT Work and Codex, including the use in the desktop app.

  2. Codex Cloud Consent form - use this form to request access to Codex Cloud, to be able to set up a Cloud environment and connect to your GitHub repository. Only current staff are able to access Codex Cloud.

Maximum monthly token usage in Codex and ChatGPT Work

Tiers

GPT-5.6 Sol

GPT-5.6 Terra

GPT-5.6 Luna

GPT-5.5

GPT-5.4

GPT-5.4-Mini

GPT-5.3-Codex

Free tier

~9.1M

~18.2M

~227.3M

~9.1M

~18.2M

~60.4M

~21.2M

Tier 2

~56.4M

~112.7M

~1.41B

~56.4M

~112.7M

~374.6M

~131.2M

Tier 3

~103.6M

~207.3M

~2.59B

~103.6M

~207.3M

~688.8M

~241.3M

Tier 4

~245.5M

~490.9M

~6.14B

~245.5M

~490.9M

~1.63B

~571.4M

Some feature consumption is measured using 'tokens'. The figures above are an estimate rather than a fixed allowance - actual limits will vary depending on workflow, prompt structure, and cache retention. It is worth noting that while the numbers in the table seem high, small tasks can consume large amounts of tokens. For example, creating an analytics dashboard based on an Excel file could use around 100,000 tokens.

The speed of the task also impacts consumption - a higher speed will allow tasks to be completed faster but at the cost of increased token consumption. If time allows, you may wish to reduce your selected speed to conserve tokens.