Safeguard
AI Security•

What a ChatGPT plan buys compared with OpenAI API prices

OpenAI publishes plan prices and estimated message counts, not token caps. Here is what its pages say and how it compares with API list prices.

Safeguard Security Team
Updated 4 min read

On 18 October 2026, OpenAI's Codex pricing page lists ChatGPT Plus at $20 a month, Go at $8, and Pro from $100. The API charges per million tokens. This post sets the two side by side using only OpenAI's own pages, and says plainly where OpenAI does not publish the number you would want.

One limit on this post: OpenAI's main chatgpt.com pricing page returned an access error when we fetched it, so the plan details below come from OpenAI's Codex pricing page at learn.chatgpt.com. That page covers coding usage across the plans, not every ChatGPT feature.

What the plans cost

PlanPrice as published
Free$0 a month
Go$8 a month
Plus$20 a month
ProFrom $100 a month, with tiers at $100, $200 and $500
Business$20 per user a month on annual billing (2 or more users), $25 monthly
Enterprise and EduContact sales

What the usage wording says

The page does not publish token caps. For Plus and standard Business plans it gives estimated local messages per five hours, and says they are "not fixed message limits" and that weekly limits may also apply:

ModelEstimated messages per five hours
GPT-6 Astra5 to 45
GPT-6.1 Sol15 to 160
GPT-6 Sol15 to 150
GPT-6 Luna350 to 3,000

The ranges are wide because message cost depends on the model, task size and complexity. The page says Pro plans currently have no five hour limit.

The API side

From OpenAI's API pricing page, Standard tier, per million tokens:

ModelInputCached inputOutput
gpt-6-astra$10.00$1.00$50.00
gpt-6.1-sol$2.00$0.10$10.00
gpt-6-luna$0.10$0.01$0.50
gpt-5.5$5.00$0.50$30.00

Batch is about half of Standard, and the page lists a Fast tier at roughly 2.5x Standard for many models. Prompts above 272,000 input tokens have their own higher rates.

Credits connect the two

The Codex pricing page also uses credits for usage beyond plan limits. Its example rates per million input tokens are 50 credits for GPT-6.1 Sol, 2.5 for GPT-6 Luna and 250 for GPT-6 Astra. Dividing those by the API input prices above gives the same ratio each time, 25 credits per API dollar. OpenAI's page does not state a dollar value per credit in the text we fetched, so treat that ratio as our arithmetic, not a published price.

Try it yourself

This snippet gives the API list price of one session on GPT-6.1 Sol. Replace the token counts with yours.

# gpt-6.1-sol Standard tier, USD per million tokens, from developers.openai.com on 18 Oct 2026
IN, CACHED, OUT = 2.00, 0.10, 10.00

def cost(fresh_in, cached_in, out):
    return (fresh_in * IN + cached_in * CACHED + out * OUT) / 1_000_000

c = cost(fresh_in=200_000, cached_in=800_000, out=20_000)
print(round(c, 2), "per session;", round(20 / c, 1), "sessions equal $20")

That tells you what $20 of API spend buys. It does not tell you what Plus includes, because OpenAI publishes message estimates, not tokens.

Quick check

Does OpenAI publish a token limit for Plus?

Not on the page we fetched. It publishes estimated messages per five hours that vary by model and are explicitly not fixed limits.

Sources

Never miss an update

Weekly insights on software supply chain security, delivered to your inbox.

Self-healing security runs on Safeguard.

Your first fix PR is minutes away.

No sales call required, even your agent can complete the purchase over MCP.