Breaking news

OpenAI Expands GPT-6 With Cheaper Sol And Luna Models As AI Pricing Pressure Intensifies

by THEFUTURE.TEAM

September 23, 2026

OpenAI is extending its GPT-6 generation with updated versions of Sol and Luna, two smaller models designed to make the company’s latest intelligence more practical, efficient and affordable for everyday business use.

OpenAI Pushes GPT-6 Down The Stack

Earlier this month, OpenAI launched GPT-6 Astra, which it described as its most powerful model yet and, in some use cases, the “world’s best model” for computer work and coding. Now, the company is broadening that platform with refreshed versions of Sol and Luna, the smaller models first introduced earlier this year.

Follow THE FUTURE on LinkedIn, Facebook, Instagram, X and Telegram

“GPT-6 Astra introduced a new generation of intelligence; these models extend its benefits by making that intelligence more efficient and accessible,” the company said.

That positioning reflects a broader strategic shift in the enterprise AI market: the race is no longer just about capability at the top end, but also about delivering those capabilities at a price and speed that make deployment scalable.

Different Models For Different Jobs

OpenAI has drawn a clear line between the two models. Sol is built for more complex work, including coding and other technically demanding tasks. Luna is aimed at higher-volume clerical use cases such as summarizing documents, extracting information and answering routine questions.

In practical terms, that mirrors how large organizations are likely to adopt AI: not as a single universal system, but as a layered stack of tools matched to specific workflows. A finance team may use one model for analysis-heavy tasks and another for document processing, just as a company would not use the same software for customer support, compliance review and software development.

Price Cuts Signal A New Phase Of Competition

The biggest commercial message in the update is not just performance. It is cost. OpenAI says the 6-series models will be available at half the price of the 5.6 versions of Sol and Luna, with the savings driven by improvements in caching and inference.

For businesses building AI products or integrating models into internal operations, that kind of reduction can matter as much as accuracy. API pricing remains one of the most important factors in enterprise adoption, especially for companies processing high volumes of queries or embedding AI into customer-facing products. A model that is slightly better but materially cheaper can quickly become the default choice.

Accuracy And Reliability Remain Central

OpenAI is also emphasizing better factual accuracy and fewer coding errors, two of the most important metrics for professional users. According to the company, GPT-6 Sol makes about half as many mistakes as its predecessor in internal factuality evaluations based on de-identified real-world conversations where users had flagged errors.

OpenAI said Sol has reached Astra-level reliability at a much lower cost. That claim underscores how the company is trying to convert raw model progress into measurable business value: not just smarter outputs, but fewer corrections, less human oversight and lower operating expense.

Anthropic Remains The Main Rival

As with previous launches, OpenAI is also using the update to sharpen its competitive message against Anthropic, arguing that GPT-6 Sol and Luna outperform the rival’s top models, including Fable and Opus.

The timing adds an extra layer of intensity. Anthropic released a new version of Opus 5.5 just 90 minutes before OpenAI’s announcement, a reminder that the frontier AI market has become a rapid-response contest in which product releases, pricing changes and performance claims are now moving in near real time.

Where The Models Are Available

The new versions of Sol and Luna are now available in ChatGPT Work and Codex for most paid accounts, as well as in the ChatGPT API. Luna will also be available in the desktop app and for Free and Go users. OpenAI said it expects to roll the models out gradually to ChatGPT’s app and website throughout the day.

For businesses, the takeaway is straightforward: OpenAI is not only advancing model capability, it is reshaping the economics of access. In a market where performance gaps are narrowing and competition is intensifying, efficiency may prove to be the feature that matters most.

The Future Forbes Realty Global Properties

Anthropic’s Opus 5.5 Arrives With Lower Costs, Faster Performance And Sharper Safety Guardrails

by THEFUTURE.TEAM

September 23, 2026

Anthropic on Tuesday unveiled Opus 5.5, its latest flagship model and, by the company’s account, a new state of the art in coding and knowledge work.

Opus remains the top tier in Anthropic’s three-model Claude family, positioned above Sonnet and Haiku, which serve the middle and entry-level segments respectively. The company says the new release not only outperforms the larger Fable model on several benchmarks, but also completed a number of informal tasks that Fable could not finish.

Follow THE FUTURE on LinkedIn, Facebook, Instagram, X and Telegram

A More Efficient Frontier Model

One of the most notable changes is economic, not just technical. Anthropic says output tokens for Opus 5.5 will be priced at $20 per million, down from $25 for the previous version. Other usage metrics have also declined, and the model is faster to run, reflecting lower compute requirements to serve it.

That matters because model economics are increasingly central to enterprise adoption. In practice, a more capable model is only part of the equation; speed and cost often determine whether it can be deployed at scale across software development, research, customer operations, and internal knowledge workflows.

Sharper Communication, Less Jargon

Anthropic says the update also changes how Opus communicates. The new model is less likely to lean on jargon and more likely to lead with the most important information first. For business users, that is more than a stylistic adjustment. It improves readability, reduces friction in decision-making, and makes AI output easier to use in executive settings where time is scarce and clarity matters.

A Rapid Follow-Up To Opus 5

The launch comes just two months after the debut of Opus 5 on July 24. Anthropic said Sonnet 5.5 and Haiku 5.5, the next models in the lineup, will follow “in the coming weeks,” with similar performance gains expected.

Safety Remains Central To The Release

Anthropic says Opus 5.5 is comparable to Mythos in biology and cybersecurity capabilities, which means the model is subject to the same safeguards as the company’s Fable model. Those restrictions limit the model’s use in areas such as discovering exploits in compiled programs or developing recognizable biological weapons, among other sensitive tasks.

The release is also notable because it is Anthropic’s first since CEO Dario Amodei publicly embraced calls to pace the frontier, a strategy designed to slow the rate of capability gains so alignment and safety measures can catch up. In a recent post, Amodei wrote: “I have become convinced that fully addressing the risks requires even more prudence, not just investing in risk prevention, but pacing the rate of capabilities advancement so that risk prevention has time to keep up.”

Preparing The Next Layer Of Oversight

Anthropic said Opus 5.5 underwent safety training broadly similar to earlier models, including alignment testing and pre-release evaluation by external groups such as METR and Frontier Design. At the same time, the company said it is already building more advanced training and evaluation systems for future releases, including stronger security and monitoring infrastructure.

“As AI becomes more capable, public policy should play a larger role in making sure the systems people rely on are safe,” the company wrote in its announcement. “That capacity takes time to build, and we’ve started to put the infrastructure in place to support it. We expect to share more details on these efforts soon.”