OpenAI Vs. Anthropic: What Their Price Cuts Mean For AI Customers

🔍 Read the full analysis: OpenAI Vs. Anthropic: What Their Price Cuts Mean For AI Customers on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

• 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 for several AI models since June, while a report by The Information says OpenAI is offering additional discounts to attract large customers from Anthropic. The changes could lower bills for businesses using AI, but public token prices do not show the full cost: custom contract discounts, usage limits and credit rules also affect what customers pay.

According to the vendors’ pricing pages, OpenAI’s Sol model line moved from $5 per million input tokens and $30 per million output tokens in June to $2 and $10 for GPT-6.1 Sol by September. Its Luna line went from $1 and $6 to 10 cents and 50 cents per million input and output tokens. The figures compare different model releases; they do not establish that the same model became cheaper while staying unchanged in capability.

Anthropic also cut prices across releases. Its Opus line moved from $5 per million input tokens and $25 per million output tokens for Opus 5 in July to $4 and $20 for Opus 5.5. Anthropic also reduced cache-read prices on some models. It says Opus 5.5 costs 40% less than Opus 5 on typical workloads, and that its Fable 5.1 model can cost about 25% less on typical workloads and up to about 45% less on highly agentic workloads. Those workload savings are Anthropic’s estimates, not independently established comparisons.

The Information reported on October 24 that OpenAI was offering customer-specific discounts. The accounts described include a reported discount of more than 30% for Replit on a contract valued at more than $100 million a year. DataCamp CEO Jonathan Cornelissen said the offer could reduce his company’s costs by 80% to 90%. That figure refers to expected overall savings against DataCamp’s current setup, according to the source material—not a discount of that size from OpenAI’s public prices.

Why AI Bills Are Harder to Compare

Lower token prices can reduce the cost of running chatbots, coding assistants and other AI services. But the public rate is only one part of the bill. Workloads vary in how many input and output tokens they use, whether cached inputs qualify for lower rates, and how much repeated context an application sends. The cost advantage of a model may differ between a short question and an agent that repeatedly processes a large conversation.

For business buyers, reported discounts raise the possibility of negotiating better terms, particularly when a company can credibly shift some work to another provider. Yet a vendor-specific offer may not be available to every customer, and its value depends on the contract, workload and service requirements. Buyers should compare actual costs on their own tasks, rather than treating a headline discount or benchmark as a forecast of savings.

Plan limits can also change without a lower monthly subscription price. The source reports that Anthropic added capped, expiring API credits to its Max and Team plans, while OpenAI reduced included usage for new $200 Pro subscriptions without changing the fee. Those changes mean customers need to review allowances and overage terms alongside advertised prices.

How the Pricing Changes Accumulated

The price moves came in a series of launches and revisions from June through September 2026. The source material says Anthropic set Claude Sonnet 5’s introductory price as its permanent price in August, then released Opus 5.5 in September at lower listed rates than Opus 5. OpenAI cut Luna prices in July and introduced GPT-6 Sol and Luna at lower rates in September, followed by a GPT-6.1 Sol cache-price change later that month.

Data scientist Max Woolf published a review of the price changes on October 25. The source author says they checked the listed figures against vendor announcements and pricing pages retrieved October 26. The sequence includes both new model releases and price adjustments, so price comparisons alone do not show whether a newer model performs equally well for a particular customer’s work.

Customer movement is also described in the reporting. Rippling CEO Parker Conrad said OpenAI’s share of the company’s coding work rose from 16% in May to 52%, while Anthropic’s share fell from 57% to 18%. Replit CEO Amjad Masad said OpenAI was its preferred partner, but later told TechCrunch that Replit remained model agnostic; a person familiar with the partnership told TechCrunch it was still in place.

“Replit remains model agnostic.”

— Amjad Masad, Replit CEO, speaking to TechCrunch

What the Headline Savings Leave Out

The full terms of the reported customer offers are not available in the supplied material. It is unclear which customers qualify, how long the discounts last, whether they depend on usage or commitments, and how the quoted contract prices compare with customers’ prior spending. The 80% to 90% estimate is attributed to DataCamp’s CEO and cannot be treated as a general saving available to other buyers.

It is also not clear from the supplied reporting how the newer models perform across customer workloads or whether reported shifts in coding use will persist. Model capability, latency, reliability and integration costs can affect total value as much as the price per token. The source reports changes to subscription allowances and credit policies, but does not provide complete plan terms for every affected tier.

What Buyers Should Check Next

Companies considering a change should test competing models on representative tasks, calculate costs using their actual token and cache usage, and compare contract terms before committing. They can ask vendors to specify discount duration, minimum commitments, usage caps, credit expiration and overage rates. Routing different tasks to more than one provider may also help buyers compare performance and retain alternatives, where their technical and contractual setup allows it.

Individual subscribers should check whether their plan’s included usage has changed, even if its monthly price has not. Further vendor pricing updates and customer contract details may clarify how broadly the reported offers apply. Until then, the confirmed public price changes and the reported private discounts should be treated separately.

Key Questions

Did OpenAI cut prices by 80% to 90%?

Not according to the figures described here. The 80% to 90% figure is DataCamp CEO Jonathan Cornelissen’s estimate of total expected savings against the company’s existing setup. It is not a stated discount from OpenAI’s public price list.

How much did the listed model prices change?

According to the source material, OpenAI’s Sol rates went from $5 input and $30 output per million tokens in June to $2 and $10 for GPT-6.1 Sol by September. Luna went from $1 and $6 to $0.10 and $0.50. These comparisons involve different model releases.

Are customers actually switching from Anthropic to OpenAI?

Some reported figures point to a shift in use at particular companies. Rippling CEO Parker Conrad said OpenAI’s share of coding work at the company rose from 16% to 52% since May, while Anthropic’s fell from 57% to 18%. That is one company’s experience, not evidence that all customers are switching.

What should businesses compare besides token prices?

They should measure performance and usage on their own workloads, then include cached-token rates, contract discounts, subscription allowances, credit expiration, seat minimums and overage charges in the comparison.

Source: ThorstenMeyerAI.com