Anthropic launches Claude Opus 5.5 with 40% lower costs and Fable 5.1-level performance
Anthropic has announced Claude Opus 5.5, a new AI model that delivers performance on par with Fable 5.1 for many tasks while reducing operational costs by 40% compared to the previous Opus 5 model. The model features improved speed, increased usage limits, and enhanced safety measures, including external evaluations by METR and Frontier Design. It is available on AWS, Google Cloud, and Microsoft…
What you can do with it
AI at Work →Draft marketing texts in one go by reading your files first
What you get You can draft emails matching landing pages and generate social media posts.
Reads files in folders, applies your tone rules, and generates drafts.
- Who for
- marketers
- Cost
- included in a tool you already have
- Effort
- minutes
Use it for
- Analyze past newsletter performance
- Draft emails matching landing pages
- Generate social media posts
Watch out Pro plan required; limit reset expires October 22
Not for everyone Pro plan only
Claude Code in the tool directory · 8 changes covered
Key points from the news
- Claude Opus 5.5 delivers Fable 5.1-level performance with 40% lower operational costs than Opus 5.
- Unit price per token decreased by 20% ($4 input, $20 output per 1M tokens vs. $5/$25 for Opus 5).
- Anthropic states Opus 5.5 outperforms GPT-6 Astra at about one-fifth the cost in GDPval-AA v2.1 evaluation.
While all sources confirm the core specifications of Opus 5.5, the note.com article provides the most detailed technical breakdown, including specific cost reductions (20% lower unit price per token and decreased token usage), real-world performance examples like migrating 680,000 lines of code in under a day, and benchmark comparisons against GPT-6 Astra. The latent.space article focuses more on broader Anthropic developments and the context of the 'Pacing the Frontier' policy, mentioning Opus 5.5's launch alongside other updates like Sonnet 5.5 and Claude Mods. The searchenginejournal.com article discusses prompting guidance for Opus 5.5, noting its default medium effort setting and recommendations for developers to adjust effort levels rather than reuse Opus 5 settings.
The reporting is consistent on the model's announcement, performance claims, cost reductions, and safety approach, with no conflicting figures or timelines across the sources.
Model pages: Claude Opus 5.5 → · Sonnet 5.5 → · GPT-6 Astra →
The story so far
9 episodes →- Anthropic launches Claude Opus 5.5 with 40% lower costs and Fable 5.1-level performancethis story
[What is Claude Opus 5.5] Fable-level performance with 40% lower costs and the highest safety to date—an explanation of what has changed
note.com · 27 September 2026
Loading the full article…
This text was published by note.com and written by 池田朋弘（いけとも）. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
19sources- Claude Opus 5.5 Beats ChatGPT 6 Astra in 8 of 12 Real-World TrialsPress · geeky-gadgets.com ·
- Anthropic's mid-tier Claude climbs the rankingsNewsletter · The Rundown AI ·
- [AINews] Opus 5.5 is good at explainer videosNewsletter · Latent Space ·
- Claude Code’s Next Era — Thariq Shihipar, AnthropicNewsletter · Latent Space ·
- How I made over 10 games with Claude Opus 5.5, including a Mario 64-style one that worked on the first tryPress · note.com ·
- Anthropic Publishes Prompting Guidance For Claude Opus 5.5 via @sejournal, @MattGSouthernPress · Search Engine Journal ·
- Claude Sonnet 5.5 is here: Comparing it with Opus 5.5 across 6 'Model x Effort' conditionsPress · note.com ·
- Claude Opus 5.5 and OpenAI Training Pause | Weekly AI News 9/21–9/28Press · note.com ·
- Results of comparing Claude Opus 5.5 and Fable 5.1: The $100/month subscription remained unchanged despite being 40% cheaperPress · note.com ·
- Why Claude Opus 5.5 is effectively 40% cheaper despite a 20% lower unit pricePress · note.com ·
- Claude Opus 5.5 Explained for Dummies in 8 MinutesPress · note.com ·
- Claude Opus 5.5 is here! If you're using AI for marketing, Claude Code is the only choicePress · note.com ·
- Technology that cuts AI costs and revision time simultaneously! How Claude Opus 5.5 is changing the 'wasted effort' in the fieldPress · note.com ·
- [Analysis] Frontier LLM Development Restrictions in Claude Opus 5.5?Press · note.com ·
- How much cheaper is Claude Opus 5.5 really? Decoding the official calculations in English (with Claude Code settings)Press · note.com ·
- Anthropic’s Claude Opus 5.5 leads Code Arena WebDev AI rankings as of Sept 2026Press · cryptobriefing.com ·
- Oil's losing streak, Trump at the UN, Royal Caribbean's Sandals stake and more in Morning SquawkPress · CNBC Technology ·
- Anthropic Launches Opus 5.5 After CEO Calls for Slower AI DevelopmentPress · en.tempo.co ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Apiiro finds 1,122 public GitHub files with secrets saved by Claude Code · 1 src
- Thinking Machines releases Inkling 2 Small, a 594B MoE model with 17B active parameters · 8 src
- OpenAI releases GPT-6 Cyber, a Critical-rated model that outperforms predecessors in vulnerability detection and alignment · 25 src
- Google releases Gemini Omni 1.2 Flash with 15-second video generation and 30% price cut to $0.07 per second · 10 src
- US accuses Chinese executive of allegedly sharing Nvidia chip info with Beijing · 11 src
Comments
via GitHub Discussions