Before Choosing An AI Subscription, Check These OpenAI And Anthropic Details

🔍 Read the full analysis: Before Choosing An AI Subscription, Check These OpenAI And Anthropic Details on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on office and shipping supplies

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

OpenAI and Anthropic have cut prices on several AI models in recent months, while OpenAI is also offering reported discounts to attract business customers from Anthropic. The developments matter to companies and individuals choosing subscriptions because the monthly price does not capture every change: usage allowances, seat minimums and credit rules can affect the actual cost.

According to vendor announcements and pricing pages reviewed by ThorstenMeyerAI.com on October 26, OpenAI’s Sol model pricing moved from $5 input and $30 output per million tokens in June to $2 and $10 for GPT-6.1 Sol in September. Its Luna line went from $1 and $6 per million tokens to 10 cents and 50 cents. These are prices for successive models, not necessarily reductions for an unchanged model.

Anthropic also lowered some prices. Its Opus line moved from $5 input and $25 output per million tokens for Opus 5 in July to $4 and $20 for Opus 5.5, while cache reads for the newer OpenAI Sol model were listed at 10 cents per million tokens. Anthropic says Opus 5.5 costs 40% less than Opus 5 on typical workloads; that comparison is the company’s estimate, not an independently established saving for every customer.

The Information reported on October 24 that OpenAI has used discounts to compete for Anthropic customers. The source article’s public summary and subsequent reporting describe an offer of more than 30% off a contract worth over $100 million a year for Replit, and a separate report of sub-$7 seats for a U.S. bank. Details of those offers have not been publicly disclosed by the companies involved.

Compare the Full Cost of AI

The price reductions give buyers more leverage in negotiations, but a lower token rate or subscription fee does not automatically mean a lower bill. Costs depend on how much a team uses a model, what work it sends to it, and whether the quoted price covers the features and capacity it needs.

Plan terms can also alter the value of a subscription. The source reports that Anthropic added capped, expiring API credits to its Max and Team plans, and that its Enterprise plan requires 20 seats. It also reports that OpenAI reduced usage included with new $200 Pro subscriptions without changing the price. Buyers should verify current terms directly, because allowances and eligibility can matter as much as the headline fee.

For organizations with substantial usage, even a negotiated discount may affect budgets and vendor choices. For individual subscribers, the main risk is paying the same monthly amount for less included access or assuming a cheaper model performs equally well on their tasks. A practical comparison should include effective usage limits, contract terms and task performance, not just list prices.

How Model Prices Shifted

The changes followed a series of model launches and pricing announcements between June and September 2026. Data scientist Max Woolf tracked the sequence in an October 25 blog post. ThorstenMeyerAI.com says it checked the figures against company announcements and pricing pages, finding them consistent. The sequence includes OpenAI price changes for Luna and Sol, Anthropic’s introductory and later pricing for Sonnet, and new Opus pricing.

Some costs that matter to AI agents are not captured by a model’s headline input and output rates. Cached tokens can be important when software repeatedly processes conversation history. Woolf estimates that cached tokens account for 96% to 99% of token spending in agentic coding; this is his estimate, not a universal measure. Anthropic has said its Fable 5.1 model costs about 25% less on typical workloads and up to about 45% less on highly agentic workloads, attributing the difference to caching. Those figures are company estimates.

Reported customer shifts add another dimension to the price competition. Rippling CEO Parker Conrad said OpenAI’s share of his company’s coding work rose from 16% in May to 52%, while Anthropic’s fell from 57% to 18%. He attributed the change to cost. These figures describe one company’s reported use and do not establish a broader market trend.

“OpenAI’s share of the company’s coding work rose from 16% in May to 52%, while Anthropic’s fell from 57% to 18%.”

— Rippling CEO Parker Conrad, as reported by The Information

Deal Terms and Usage Remain Unclear

The publicly available reporting does not provide the full terms of the reported customer offers, including contract duration, eligibility, usage commitments or which services the discounts cover. The Information’s article is paywalled, and some details in the supplied account come through secondary reporting or people familiar with the arrangements.

It is also not clear how broadly the reported discounts are available or whether they apply to other customers. The sub-$7 bank seats and the reported Replit discount are specific deals, not published standard rates. Replit’s relationship with Anthropic has also been described differently: Masad said the company was no longer working with Anthropic, while a person familiar with the partnership told TechCrunch it remained in place.

Finally, successor-model prices do not establish that a newer model will perform as well as, or better than, an older one for every job. Buyers need to test models on their own workloads and confirm current plan allowances with vendors before making a decision.

Verify Terms Before Renewing

Companies considering a renewal or vendor change can use the price competition to request current written quotes and ask vendors to specify rates, included usage, overage charges, seat minimums and credit expiry rules. Where practical, testing more than one provider on representative tasks can reveal whether a lower-priced model meets the team’s quality and reliability needs.

Individual subscribers should check the latest usage allowance and restrictions for the specific plan, rather than relying on an earlier description or on the monthly price alone. The reported changes point to continuing competition, but the next public price move or plan revision is not specified in the source material. Buyers will need to monitor vendor pricing pages and contract notices for updates.

Key Questions

Does the reported 80% to 90% saving mean OpenAI cut its list prices by that amount?

No. The range refers to DataCamp CEO Jonathan Cornelissen’s reported estimate of total savings against the company’s existing setup. It is not a general discount from OpenAI’s published prices.

Which OpenAI prices changed?

According to the source’s review of vendor pricing pages, Sol moved from $5 input and $30 output per million tokens in June to $2 and $10 for GPT-6.1 Sol in September. Luna moved from $1 and $6 to 10 cents and 50 cents. The figures concern successive models.

What should I check before buying an AI subscription?

Check the plan’s included usage, overage charges, seat minimums, credit limits and expiry dates, as well as the listed monthly price. Test the model on the tasks you actually need it to perform.

Are OpenAI’s reported enterprise discounts available to everyone?

The source material describes specific customer offers, including one reported Replit discount and a bank seat price. It does not establish that those terms are publicly available or offered to all customers.

Source: ThorstenMeyerAI.com