AI News Feed
Market watch
Enthusiast rebuilds M4 Mac Mini into a 3D-printed handheld with a 7-inch 120Hz screen, keyboard and batteryCisco Talos Details Windows Malware That Lets Four LLMs Vote on Its Next MoveCognex to buy Intel robotics spin-off RealSense for about $600MEngadget Outlines Ways to Improve Apple CarPlay as iOS 27 Adds FeaturesTechRadar Publishes Sept. 23 Answers for Quordle, NYT Strands and ConnectionsApple Releases iOS 27.2 and macOS 27.2 Public Betas, Second RCs for Tahoe and Sequoia
Enthusiast rebuilds M4 Mac Mini into a 3D-printed handheld with a 7-inch 120Hz screen, keyboard and batteryCisco Talos Details Windows Malware That Lets Four LLMs Vote on Its Next MoveCognex to buy Intel robotics spin-off RealSense for about $600MEngadget Outlines Ways to Improve Apple CarPlay as iOS 27 Adds FeaturesTechRadar Publishes Sept. 23 Answers for Quordle, NYT Strands and ConnectionsApple Releases iOS 27.2 and macOS 27.2 Public Betas, Second RCs for Tahoe and Sequoia
Large Language Models

Anthropic Releases Claude Opus 5.5 as OpenAI Counters With Cheaper GPT-6 Sol and Luna

Anthropic released Claude Opus 5.5 on Sept. 22 with a 20% price cut, and OpenAI minutes later launched GPT-6 Sol and Luna at half the cost of their predecessors, according to SiliconANGLE.

AI News Feed · 2026-09-23 09:09

Anthropic priced Opus 5.5 at $4 per million input tokens and $20 per million output tokens. The company said that works out to 40% less on a typical workload than Opus 5. Cache reads took the sharpest reduction, falling 60% to 20 cents. A faster serving mode costs $8 per million input tokens and $40 per million output tokens.

OpenAI halved the price of both new models against the GPT-5.6 versions that carried the same names. Sol costs $2 per million input tokens and $10 per million output tokens. Luna is priced an order of magnitude lower, at 10 cents per million input tokens and 50 cents per million output tokens.

Anthropic said Opus 5.5 matches its Fable 5.1 model on most work and generates output more than 30% faster than Opus 5, which shipped in late July. On Terminal-Bench 4.0, an agentic coding test, Opus 5.5 scored 66.4%, compared with 55.8% for Fable 5.1. On AutomationBench, it reported 40% task completion, while Opus 5 managed 26.9%.

The widest gap in the published figures was in scientific research. On Terminal-Bench-Science 0.1, Opus 5 scored 29% and Opus 5.5 scored 58.7%. On GDPval-AA v2.1, a knowledge work test scored in Elo, Opus 5.5 rated 1846 against 1708 for Opus 5. Humanity’s Last Exam, run with tools, returned 67.7%. One tester finished a 680,000-line code migration in under a day, work Anthropic said would have taken an engineering team weeks.

In testing notes Anthropic published with the release, Yashodha Bhavnani, vice president of AI products at Box Inc., said answers from the new model were 40% less verbose without losing accuracy. GitHub Inc. Chief Product Officer Mario Rodriguez said that in VS Code the model “solved more terminal tasks than Opus 5 in less than half the steps.” Anthropic said the model puts the most important information up front. Other customers quoted in the announcement described fewer retries and less rework.

On safety, Anthropic called Opus 5.5 the strongest performer to date on the automated behavioral audit it uses for alignment testing. It said attempts to circumvent containment boundaries dropped 85% against Opus 5, and resistance to prompt injection improved enough to match Fable 5.1 on the Gray Swan benchmark. Restrictions on cybersecurity work carry over from Fable 5.1, with most such tasks routed to the older Opus 4.8 model. High-risk biology work is fenced off as well, and broader access runs through verification programs that now include a life sciences track. Frontier Design and METR both tested the model before release.

OpenAI aimed its two releases lower down the cost curve. Sol and Luna were built on much the same training methods as GPT-6 Astra, which the company began rolling out Sept. 3, then tuned for cost. Luna is the cheaper of the pair and is pointed at high-volume routine jobs such as summarization and extraction. Sol takes on recurring coding and agent work. Astra for Law, a configuration built for legal research, arrived Sept. 17.

OpenAI reported Sol completing 33.2% of tasks on AutomationBench at 27 cents apiece and put it at 68.8% on the DeepSWE v1.1 software engineering test, within 1.1 points of the earlier Claude Fable 5. Luna reached 66.6% on the same test. Sol makes about half as many mistakes as its predecessor, according to the company, and Luna at higher effort settings matches GPT-5.6 Sol at roughly a hundredth of the cost. OpenAI’s comparisons run against Fable 5.1 and Opus 5, so neither launch offers a direct head-to-head against the other.

OpenAI is discounting reads of cached input tokens by 90%. It said hit rates have improved, and that coding agents can now change reasoning effort and the set of tools available to them without breaking the cache. Answers from both models come back shorter, with less jargon and fewer low-value details, by the company’s account.

Opus 5.5 is available on the Claude Developer Platform as claude-opus-5-5 and through Amazon Web Services Inc., Google Cloud and Microsoft Azure. Claude Sonnet 5.5 and Claude Haiku 5.5 are due in the coming weeks. The OpenAI models are live for developers as gpt-6-sol and gpt-6-luna, and in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu subscribers. Free and Go users get Luna in the desktop app, and neither model has reached the main ChatGPT chat surface yet.

Editor's Summary

Anthropic released Claude Opus 5.5 with a 20% price cut and improved benchmarks, while OpenAI quickly answered with GPT-6 Sol and Luna at half the cost of earlier GPT-5.6 models. The back-to-back launches continued a push toward cheaper inference and more capable agentic and coding models. Availability and safety restrictions differed, with Anthropic limiting some cybersecurity and biology work and OpenAI positioning Sol and Luna for cost-sensitive developer and workplace uses.