🔍 Read the full analysis: Should You Switch AI Providers? Compare OpenAI And Anthropic’s Offers on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI and Anthropic have reduced prices for several models, and reporting describes OpenAI offering discounts to win business from Anthropic. The reported 80–90% savings figure is a customer’s estimate of total cost reductions, not a general discount off OpenAI’s list prices; buyers also need to check usage limits, credits and seat minimums.
OpenAI and Anthropic have lowered prices for several AI models in recent months, while OpenAI is also offering customer-specific discounts to compete for business, according to reporting by The Information. The changes may give companies more leverage to negotiate, but the widely cited 80–90% savings figure describes one customer’s expected overall cost reduction, not a standard discount on OpenAI’s published prices.
Published pricing pages reviewed by ThorstenMeyerAI.com on October 26 show a substantial drop in listed token rates for some models. 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 in September. Its Luna line fell from $1 and $6 to $0.10 and $0.50. These comparisons involve successor models, so they do not establish that the same model became cheaper while remaining identical in capability.
Anthropic also reduced rates. Its Opus pricing moved from $5 per million input tokens and $25 per million output tokens for Opus 5 to $4 and $20 for Opus 5.5. The company also cut prices for some cached inputs. Anthropic says its newer models can cost less on typical workloads, but those workload savings are company estimates, not independently verified results for every customer.
The Information reported on October 24 that OpenAI is using discounts to attract customers associated with Anthropic. The reported examples vary: one person cited in the report said Replit received an offer of more than 30% off a contract worth over $100 million a year. DataCamp CEO Jonathan Cornelissen said he expects OpenAI’s pricing to reduce his company’s costs by 80% to 90%. That figure refers to expected savings against its existing setup, rather than a discount from OpenAI’s list price.
Infographic 01 · The price slide
Four months, a fraction of the price.
What a million output tokens cost on each vendor’s price list, from the June previews to today. Bars show the price; shorter is cheaper.
OpenAI · output price per 1M tokens
List prices, standard tier
Anthropic · output price per 1M tokens
List prices, standard tier
Same scale in both panels. Fable 5.1 kept its $50 output price and cut cache reads instead.
The quieter cut: cache reads
matters most for agents
Cache-read price per 1M tokens, same scale in all three.
Woolf estimates cached tokens make up 96–99% of token spend in agentic coding. Anthropic says Opus 5.5 costs 40% less than Opus 5 and Fable 5.1 about 25% (up to about 45%) less than Fable 5 on typical workloads; those are vendor claims. Most rows compare a successor model with its predecessor. Anthropic’s newer tokenizer produces about 30% more tokens for the same text, so per-token prices are not directly comparable across vendors.
Sources: OpenAI API pricing (26 Oct 2026); OpenAI announcements (26 Jun–29 Sep 2026); Anthropic API pricing (26 Oct 2026); Anthropic announcements (30 Jun–22 Sep 2026); Vercel AI Gateway index (Oct 2026); Max Woolf (25 Oct 2026).
Powered by Thorsten Meyer AI
Lower Rates, More Negotiating Leverage
For businesses with substantial AI usage, changes in per-token pricing can affect operating budgets, especially when coding agents or other tools make repeated model calls. The price quoted on a vendor page, however, is only one part of the bill. Usage allowances, cached-token rates, seat minimums, credit expiry and contract-specific discounts can all affect what an organization pays in practice.
Customer movement suggests that the competition is affecting deployment choices, though the available examples do not establish a market-wide shift. Rippling CEO Parker Conrad said OpenAI’s share of coding work at his company rose from 16% in May to 52%, while Anthropic’s share fell from 57% to 18%. Those figures describe one company’s usage, not a general market share measure. They show why businesses may want to compare performance and cost on their own workloads rather than rely on headline prices.
For smaller customers and individual subscribers, a lower model price does not necessarily mean more included usage. The source material says OpenAI reduced the usage included in new $200-per-month Pro subscriptions without changing the monthly fee. Anthropic added capped, expiring API credits to some Max and Team plans. A buyer comparing monthly charges alone could miss limits that shape the actual service received.
Products Worth Considering
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
LG UltraGear 45GX900A-B 45-Inch Curved OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How the Price Cuts Accumulated
The reductions followed a series of launches and pricing changes from June through September 2026. Max Woolf, a data scientist who tracks language-model pricing, published a timeline on October 25. ThorstenMeyerAI.com says it checked the listed figures against vendor announcements and current pricing pages retrieved the following day. The sequence included OpenAI price cuts, Anthropic’s decision to make introductory Sonnet pricing permanent, and changes to cache rates.
Those distinctions matter because some apparent savings came with new model versions, while others involved promotional prices or specific token types. Anthropic attributed lower costs for certain workloads to cached inputs, which can make up a large share of usage in agentic coding. The source cites Woolf’s estimate that cached tokens account for 96% to 99% of token spending in such work. That estimate is not a universal measure for all AI applications, and cache economics will vary by workload.
Contract terms are also changing alongside prices. The source says Anthropic’s Enterprise plan requires a 20-seat minimum, potentially limiting access for smaller firms. It also reports that some Anthropic plans now include credits with caps and expiry dates. These conditions are separate from the per-token rates and should be checked against the exact plan or contract a customer is considering.
“I think ChatGPT is better now.”
— Replit CEO Amjad Masad, speaking at TechCrunch Disrupt on October 16
Contract Savings Still Vary
The public information does not provide a complete set of contract terms or enough detail to independently calculate the reported customer savings. The Information’s article is paywalled; the source material says its account was reviewed alongside public summaries and other reporting. It is not clear what workloads, usage volumes, service commitments or existing vendor rates underpin DataCamp’s 80% to 90% estimate.
Nor do the listed prices settle which model is the better value for a particular task. The compared products are often different generations, and vendor claims about relative intelligence or workload savings have not been independently established here. Replit’s relationship has also been described differently: Masad called OpenAI its preferred partner and said it remained model agnostic, while a person familiar with the partnership told TechCrunch it was still in place. The precise commercial arrangement and how it may change are not specified.
Some plan details may change after the October 26 review, and pricing pages do not capture every negotiated enterprise contract. The source does not provide complete terms for the reported U.S. bank offer, including its exact price, duration or conditions.
Compare Workloads and Contract Terms
Companies weighing a switch can compare vendors using representative tasks and actual usage data, then review the full proposed contract rather than extrapolating from a list price. That comparison should include model quality for the organization’s work, input and output token rates, cached-token pricing, usage caps, credit expiry, seat requirements and any service commitments. A negotiated offer should also be checked for its duration and what happens when promotional terms end.
The immediate next step for buyers is to request current written terms from each provider and test likely workloads before committing. The price changes and reported offers show active competition, but the information available does not establish that either provider will be cheaper or better for every customer. Future vendor pricing updates and contract disclosures may clarify how durable the discounts are and whether the reported shifts in customer usage continue.
Key Questions
Did OpenAI offer everyone a 90% discount?
No. The 80% to 90% figure was DataCamp CEO Jonathan Cornelissen’s estimate of expected total savings against the company’s existing setup. It was not reported as a general discount off OpenAI’s published prices.
How much did the listed model prices fall?
In the figures reviewed by ThorstenMeyerAI.com, OpenAI’s Sol rate moved from $5 per million input tokens and $30 per million output tokens in June to $2 and $10 for GPT-6.1 Sol in September. These are different model versions, so the comparison does not show an identical model becoming cheaper.
Are businesses moving from Anthropic to OpenAI?
Some reported examples point to changes in customer usage and offers, but they do not establish a broad market trend. Rippling’s CEO reported a change in the share of coding work handled by each provider at his company; other firms’ choices and contract details may differ.
What should buyers compare besides token prices?
Review usage limits, cached-token rates, seat minimums, credit caps and expiry, contract duration, and any service commitments. Test models on representative tasks because model capability and total cost depend on the workload.
Can customers assume their subscription includes the same usage as before?
No. The source reports that OpenAI reduced usage included with new $200-per-month Pro subscriptions without changing the monthly price, while Anthropic added capped, expiring credits to some plans. Customers should check the current terms for their specific plan.
Source: ThorstenMeyerAI.com
