Anthropic发布Opus 5.5大模型，多项基准超越Fable且价格下调
人工智能公司Anthropic于2026年9月22日发布新一代模型Opus 5.5。据公司介绍，该模型在编程和知识工作基准测试中创下新纪录，多项测试成绩超越更大的Fable模型，并完成了一些Fable未能完成的非正式任务。
价格方面，Opus 5.5的输出token定价从上一代的每百万25美元降至20美元，其他计费指标也有类似下调。Anthropic表示，模型的运行速度更快，反映出服务所需算力整体下降。
新版模型在交互方式上也有所调整，Opus 5.5更少使用专业术语，更倾向将重要信息放在回复开头。此次发布距离Opus 5于2026年7月24日推出仅约两个月。官方公告称，Sonnet 5.5和Haiku 5.5将在未来数周内发布，并带来类似的性能提升。
Anthropic称，Opus 5.5在生物学和网络安全方面的能力与Mythos相当，因此发布过程采用了与Fable模型相同的安全保护措施，限制模型在编译程序漏洞利用或开发可识别生物武器等任务上的使用。
这是Anthropic首席执行官Dario Amodei提出“为前沿进展设定节奏”以来公司的首次模型发布。Amodei本月早些时候发文表示，必须采取更多审慎措施，不仅投资于风险防范，还要调整能力发展速度，使风险防范有时间跟上。
在安全训练方面，Opus 5.5与此前版本大体类似，接受了对齐测试以及METR、Frontier Design等外部机构的发布前评估。Anthropic同时强调，正在为未来模型准备更先进的训练与评估体系，包括改进的安全和监控系统，并预计在后续公布更多细节。
- 2026-09-24 08:23 | 界面新闻：AI早报 | 马斯克预测20年后可能有1000亿台人形机器人；Anthropic发布Claude Opus 5.5阅读原文
马斯克预测未来世界：20年后可能有1000亿台人形机器人 9月23日，特斯拉CEO埃隆・马斯克接受央视财经专访。谈及机器人，他展望未来美好图景：每个人都能拥有专属机器人，它们可以照料年迈的父母，看护孩童，还能担任孩子的一对一私教，承担各类工作。未来世界，一个人可以操控成百上千台，甚至上万台实体机器人和数字智能体。“我的预测是十年内至少有10亿台人形机器人；未来15年或许达到100亿台；未来20年可能达到1000亿台。” Anthropic发布Claude Opus 5.5 当地时间9月22日，Anthropic宣布推出Claude Opus 5.5，该模型在大多数工作中都能达到Claud...
- 2026-09-23 10:14 | 澎湃新闻：AI“减速”讨论未歇，OpenAI、Anthropic同日上新，竞逐更低成本阅读原文
安全讨论持续的同时，人工智能巨头争夺用户与应用市场的竞争仍在继续。当地时间9月22日，OpenAI发布GPT-6 Sol和GPT-6 Luna，宣布两款模型的API价格较上一代GPT-5.6对应模型的促销定价下调50%。同时，Anthropic也正式推出了Claude Opus 5.5，称其在大多数工作中的表现可达到Claude Fable 5.1的水平，典型任务的运行成本则较Opus 5降低40%。OpenAI、Anthropic同日上新，竞逐更低成本。图片由AI生成此次OpenAI发布的两款模型，是对GPT-6产品系列的进一步扩充。此前，OpenAI已推出GPT-6 Astra。OpenA...
- 2026-09-23 03:15 | gizmodo.com：Ten Days After CEO Calls for a Slowdown, Anthropic Is Back With Another AI Model阅读原文
Less than two weeks after Anthropic CEO Dario Amodei urged the AI industry to “pace the frontier” before it loses control of the technology it’s currently racing to build, the company has released its most powerful model to date. Unveiled Tuesday, Claude Opus 5.5—the first in a new family of models,...
- 2026-09-23 03:05 | venturebeat.com：Anthropic releases Claude Opus 5.5, beating Fable 5.1 on key agentic benchmarks at 60% cheaper API price阅读原文
Anthropic today released Claude Opus 5.5, its new frontier model aimed at long-running coding agents, research, and professional knowledge work and it both outperforms and undercuts the cost (over application programming interface, or API) of Anthropic's prior flagship models Mythos 5.1 and Fable 5....
- 2026-09-23 00:44 | X：Per Anthropic, Opus 5.5 costs 40% less than Opus 5 to run on typical workloads at default settings a...阅读原文
Per Anthropic, Opus 5.5 costs 40% less than Opus 5 to run on typical workloads at default settings and generates output over 30% faster. Their full announcement:Claude@claudeaiIntroducing Claude Opus 5.5, the first model in our new Claude 5.5 family.It performs at the level of Claude Fable 5.1 for m...
- 2026-09-23 00:44 | X：Claude Opus 5.5 from @AnthropicAI is live on OpenRouter! The first model in the Claude 5.5 family l...阅读原文
Claude Opus 5.5 from @AnthropicAI is live on OpenRouter!The first model in the Claude 5.5 family leads Opus 5 and Fable 5.1 on agentic coding, knowledge work, and computer use, with 1M context at $4/M input and $20/M output, 20% lower per token than Opus 5.
- 2026-09-23 00:44 | X：Re Opus 5.5 is available now in Claude Code and the Claude Platform. With its intelligence, lower price, and speed, it's our new daily driver for codi...阅读原文
Re Opus 5.5 is available now in Claude Code and the Claude Platform. With its intelligence, lower price, and speed, it's our new daily driver for coding and agentic work. We hope it'll become yours too.
- 2026-09-23 00:44 | X：Re On the Claude Platform, Opus 5.5 is $4/$20 per million tokens, down from $5/$25. Cache reads drop 60%, from $0.50 to $0.20. Run `/claude-api migrat...阅读原文
Re On the Claude Platform, Opus 5.5 is $4/$20 per million tokens, down from $5/$25. Cache reads drop 60%, from $0.50 to $0.20. Run `/claude-api migrate` in Claude Code to switch to `claude-opus-5-5`. Our guide to lower costs with the Claude Platform: https://claude.com/blog/reducing-cost-and-improvi...
- 2026-09-23 00:44 | X：Opus 5.5 performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task. In Claude Code: - 5-hour session limits increase ...阅读原文
Opus 5.5 performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task. In Claude Code: - 5-hour session limits increase 20% today - Opus 5.5 is priced lower, so it goes 25% further within limits - Pro, Max, and Team users get a reset to use anytime Claude: Introducing ...
- 2026-09-23 00:31 | X：Claude Opus 5.5 is available today.阅读原文
Claude: Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
- 2026-09-23 00:30 | TechCrunch：Anthropic releases Opus 5.5 with lower prices and Fable-level performance阅读原文
Anthropic’s newest model, Opus 5.5, was released on Tuesday, setting a new state-of-the-art in coding and knowledge work performance, according to the company. Notably, the company says, the release outpaces the larger Fable model in many benchmarks, and succeeded in a number of informal tasks that ...
- 2026-09-23 00:30 | zdnet.com：Claude Opus 5.5 delivers Fable 5.1 performance – and costs 40% less阅读原文
Elyse Betters Picaro/ZDNETZDNET’s key takeaways Claude Opus 5.5 could be a big win for power users. Developers may see faster coding with fewer steps. Anthropic says the upgrade is safer, cheaper, and less wordy. Less than two months after the release of Claude Opus 5, Anthropic is back wit...