Anthropic launches Claude Opus 5.5 with enhanced safeguards
Anthropic released Claude Opus 5.5, its first model since CEO Dario Amodei's 'pace the frontier' essay, with enhanced safeguards against risky behaviors such as attempts to escape testing environments. The model tops multiple benchmarks, with output speed over 30% faster than Opus 5, API prices cut by 20%, and cache-read prices reduced by 60%.
Coverage timeline
The Verge AIEmma Roth
Anthropic says its new Claude Opus 5.5 model comes with stronger safeguards in the wake of recent rogue AI hacking incidents. In an announcement on Tuesday , Anthropic says Opus 5.5 comes with improvements to certain risky behaviors, including attempts to escape the company's testing sandbox. It's the first model released by Anthropic after CEO Dario Amodei announced plans to "pace the frontier," or slow down AI development. In recent weeks, several AI companies, including Anthropic , Google , and OpenAI , have reported that their AI models escaped containment and hacked third-party companies during testing. Anthropic says Opus 5.5 is the "str … Read the full story at The Verge.
Techmeme
Emma Roth / The Verge : Anthropic launches Claude Opus 5.5, its first model since Dario Amodei's “pace the frontier” essay, and says it has enhanced safeguards to combat risky behavior — Claude Opus 5.5 comes with improvements to certain behaviors, like attempting to escape testing environments.
Simon Willison
Release: llm-anthropic 0.29 Adds support for Claude Opus 5.5 : llm -m claude-opus-5.5 "prompt goes here" Tags: llm , anthropic
量子位量子位
Claude Opus 5.5突袭！68万行代码一天迁完，API价格打8折 听雨 2026-09-23 12:59:29 来源： 量子位 缓存读取价砍了60% 梦晨 听雨 发自 凹非寺 量子位 | 公众号QbitAI Claude最新旗舰 Opus 5.5 跑分登顶！ 在代码、知识工作、计算机操作等多项基准测试中拿下第一的同时，输出速度也比 Opus 5 快了 30%以上 。 这暂停得好好的前沿，怎么不到两周又被推进了。 那CEO Dario Amodei发表的“放缓前沿”公开信算什么？ 算他能发。 这届模型开始拿代码当画笔 目前看到比较新颖的测试是用纯代码生成图像，SVG鹈鹕骑自行车已经被刷爆了，这次是更复杂的Python渲染。 类似的能力可以扩展成用纯js代码生成渲染视频。 多项跑分登顶，API价格打8折 他们说任务成本降 4成 ，但API价格并没有直接打 6折 。 Opus 5.5的输入、输出价格分别是每百万Token 4美元 和 20美元 ，相比Opus 5下降 20% 。 再加上完成同一任务需要的步骤和Token更少，Anthropic测算，典型任务的总成本可以下降 40% 。 真正下狠手的是 缓存读取价 。这一项从Opus 5的每百万Token 0.50美元 ，直接降到 0.20美元 ，砍了 60% 。 做Agent编程的都知道，长对话和大型代码任务需要反复读取上下文，缓存读取经常占据成本的大头。相比账面上的API单价，这一刀可能更有实际体感。 另外还有一个Fast mode，输出速度最高可达标准模式的 2.5倍 ，价格也翻倍至每百万输入Token 8美元 、输出Token 40美元 ，适合对延迟特别敏感的场景。 订阅用户也加量了，Pro、Max、Team和按席位收费的企业版，五小时使用限额都会提高；Anthropic还提供了一次可以暂存、由用户自行选择时间使用的rate limit reset。 具体看跑分，目前最新鲜的榜就是 Terminal-Bench 4.0 了，衡量模型处理复杂命令行任务的能力。 Opus 5.5拿下 66.4% ，上一代Opus 5是 52.3% ，OpenAI的 GPT-6 Astra 是 57.9% ，Claude自家的 Fable 5.1 是 55.8% 。 已经明显拉开差距。 在另一项编程基准FrontierCode
Related stories
Anthropic: Claude Traces NIST ECDSA Curve Seeds to PRNG, Easing Backdoor Fears
Anthropic reported that its Claude research model reverse-engineered the generation of five NIST elliptic curve seeds in under half an hour, without prior hints. The seeds were traced to the lrand48 pseudorandom number generator, seeded by a Unix timestamp, challenging decades-old suspicions of an NSA backdoor. The finding was independently verified by the bounty sponsor and lowers concerns about hidden weaknesses in the curves.
4 reports · 4 sources1058 points · 388 comments
Claude Code team shifts code review to continuous AI-assisted checks
Members of Anthropic's Claude Code team described how they rethought code review, moving from a final human gate to continuous AI-assisted checks with Claude before human review. They argue that 'did a human read the code?' is an imprecise proxy for identifying potential issues, and that most changes are now reviewed and fixed by Claude at least once before human scrutiny, freeing attention for higher-level questions.
Anthropic: New measures cut distillation-attack data harvesting by 90-99%
Anthropic reports that new measures, including preserved thinking in Fable 5.1 and Opus 5.5, have reduced data harvesting through distillation attacks on Claude by approximately 90% to 99%. The company detailed these methods in a recent blog post, marking substantial progress in mitigating such attacks.
1 reports · 1 sources
Anthropic launches OSS Scanner, a free opt-in vulnerability scanner for open-source projects
Anthropic launched OSS Scanner, a free opt-in vulnerability scanner for critical open-source projects, on October 8, 2026. The service uses Anthropic's strongest models to perform periodic security scans and generate reports without human review or triage, aiming to enable faster and more frequent scanning while acknowledging potential inaccuracies.
3 reports · 3 sources
News updates every hour. Subscribe in your RSS reader to get new stories automatically.
Subscribe