🔍 Read the full analysis: OpenAI Vs. Anthropic: A Closer Look At Their AI Price Competition on ThorstenMeyerAI.com
Get the latest gadgets 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 reduced prices for several AI models in 2026, while reports describe OpenAI offering discounts to win enterprise customers. A reported 80% to 90% savings figure refers to one buyer’s expected total cost reduction, not a general discount off OpenAI’s list prices. Contract terms, usage limits and model performance remain part of the comparison.
OpenAI and Anthropic have cut prices for several AI models since June, and a report published October 24 says OpenAI is also offering customer-specific discounts to win business from Anthropic. The changes give companies more leverage to negotiate, but headline token prices do not show the full cost: usage limits, credits, caching and contract terms can also affect what customers pay.
According to the vendors’ pricing pages, OpenAI’s GPT-6.1 Sol costs $2 per million input tokens and $10 per million output tokens, down from $5 and $30 for the earlier Sol price listed in June. Its Luna model fell from $1 and $6 per million input and output tokens to $0.10 and $0.50. These are prices for successive models, not simply a reduction for an unchanged model.
Anthropic also lowered prices across its lineup. Its Opus pricing moved from $5/$25 to $4/$20 per million input/output tokens, while the source reports that cache reads for a later model were cut by 60%. Anthropic has said Opus 5.5 costs 40% less than Opus 5 on typical workloads and that its Fable 5.1 model costs around 25% less on typical workloads. Those workload savings are the company’s estimates, not independently established results.
The reported customer offers vary. The Information reported a discount of more than 30% on a Replit contract valued at more than $100 million a year, according to a person cited in its reporting. DataCamp CEO Jonathan Cornelissen described expected overall cost savings of 80% to 90% from OpenAI’s offer. That figure is not a 90% reduction in OpenAI’s public list price; it is an estimate of savings against DataCamp’s current setup.
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
How Lower Prices Change Buyer Leverage
For companies buying AI services, falling list prices and reports of negotiated discounts make it more valuable to compare offers and revisit existing contracts. A lower per-token rate can reduce bills, but only if the model performs adequately on the tasks at hand and the deal’s usage rules fit the customer’s needs. Total cost depends on more than list price, including how much work is routed to a model and how cached inputs are charged.
The reported changes in customer use suggest price and product performance can affect vendor choice. Rippling CEO Parker Conrad said OpenAI’s share of his company’s coding work rose from 16% in May to 52%, while Anthropic’s share fell from 57% to 18%. These figures are his account of one company’s usage, not a measure of the wider market. Still, they illustrate why suppliers are competing over both pricing and customer workloads.
Individual users also face a changing value proposition. Some plans may offer access to cheaper models, but monthly subscription prices alone may not reveal whether the included usage has changed. Allowances and plan limits matter alongside the advertised fee.
Products Worth Considering

LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDR True Black 400

Kindle Paperwhite Signature Edition 32GB (2024 Model) – Metallic Black

LG 27GR83Q-B 27-inch Ultragear QHD IPS Gaming Monitor
As an affiliate, we earn on qualifying purchases.
A Four-Month Run of Price Changes
The source material, citing data scientist Max Woolf’s October 25 pricing review and vendor announcements and pricing pages checked October 26, describes a series of changes from June through September. OpenAI and Anthropic introduced new models, adjusted prices and changed cache pricing. Anthropic made an introductory Sonnet price permanent, while OpenAI cut Luna pricing and later priced its GPT-6 models below earlier listed levels.
Some comparisons involve successor models rather than the same model becoming cheaper. Customers therefore need to test whether a lower-priced replacement produces results that work for their own tasks. Cache pricing can also alter the bill for coding agents that repeatedly send conversation history: the source cites Woolf’s estimate that cached tokens can account for 96% to 99% of token spending in agentic coding. That estimate applies to the described workload, not all AI use.
Reported business terms are separate from public pricing. The source says Anthropic added capped, expiring API credits to Max and Team plans and requires 20 seats for Enterprise. It also says OpenAI reduced usage included with new $200 Pro subscriptions without changing the subscription price. These plan details can change the value of an offer even when its advertised price stays constant.
What the Reported Discounts Cover
The public information does not establish how widely OpenAI is offering customer-specific discounts, which contract terms apply, or how long the offers last. The 80% to 90% figure is one executive’s estimate of expected savings; it is not a verified general rate. The source’s account of the Replit discount relies on a person cited by The Information, and the precise terms have not been provided publicly.
There is also uncertainty around how comparable model prices are in practice. Successor models may behave differently on a customer’s tasks, and vendor estimates of workload savings depend on the workloads and assumptions used. Plan allowances and credit rules can change the effective price, while a lower token rate may not lead to a lower bill if usage rises. Public pricing alone cannot settle those comparisons.
What Buyers Should Compare Next
Companies considering a renewal or a switch will need to compare vendor proposals against their actual usage, including token volume, cache reads, seats, included allowances, credit expiry and any minimum commitment. Testing candidate models on representative tasks can help buyers distinguish a lower price from a lower-cost service that still meets their needs. The source recommends negotiating with vendors and routing work across more than one provider where practical.
Further price or plan changes may follow, but the source does not identify a scheduled announcement or next milestone from either company. Buyers should monitor updated pricing pages and contract terms rather than assume current discounts or usage limits will remain in place.
Key Questions
Did OpenAI cut its prices by 90%?
No general 90% list-price cut is established. The 80% to 90% figure refers to DataCamp CEO Jonathan Cornelissen’s expected total savings against the company’s existing setup. Separately, the source reports public price reductions for specific OpenAI models.
How much did OpenAI’s listed token prices change?
The source says GPT-6.1 Sol is listed at $2 per million input tokens and $10 per million output tokens, compared with $5 and $30 for the earlier June Sol price. Luna moved from $1/$6 to $0.10/$0.50 per million input/output tokens. These comparisons involve different model versions.
Did Anthropic also lower prices?
Yes. The source reports that Opus pricing moved from $5/$25 to $4/$20 per million input/output tokens, and that Anthropic cut cache-read prices for a later model. Some savings claims are Anthropic’s estimates for typical workloads.
Why do plan terms matter if token prices are falling?
Usage allowances, seat minimums, capped or expiring credits, and cache pricing can affect the final bill. A monthly fee or per-token rate by itself may not show how much service a customer receives.
Source: ThorstenMeyerAI.com
