AIToday
Large Language ModelsAI Coding AssistantsAI Business & IndustryTechCrunch AIPublished: Sep 23, 2026, 04:00 JST

Anthropic releases Opus 5.5 at $20, beating Fable

Anthropic releases Opus 5.5 at $20, beating Fable

3 Key Points

  1. What happened

    Anthropic released Opus 5.5 on Tuesday, saying it sets a new state-of-the-art in coding and knowledge work, beats Fable in many benchmarks, and charges $20 per million output tokens versus $25 before.

  2. Why it matters

    The release is both cheaper and faster to run, reflecting a drop in the compute required to serve it.

  3. What to watch

    The launch comes just two months after Opus 5 on July 24, and Anthropic says the next tiers, Sonnet 5.5 and Haiku 5.5, will follow 'in the coming weeks'.

WHO IT HITSEnterprise teams choosing which Anthropic model tier to run for coding and knowledge-work tasks may see lower output costs and faster runs. Developers subject to Anthropic's Fable-level safeguards for biology and cybersecurity will face the same usage limits on Opus 5.5.

Not sure about something? Ask the AI

Summaries like this, in your inbox every morning.

Context & Analysis

Anthropic's Opus line sits at the top of its three-tier lineup, with Sonnet in the middle and Haiku as the fastest and cheapest option. The company says Opus 5.5 outpaces the larger Fable model in many benchmarks and completed several informal tasks that Fable failed. The previous Opus 5 launched on July 24, making this a two-month cadence between top-tier releases.

The price cut to $20 per million output tokens, down from $25, and the faster run time reflect an overall drop in the compute required to serve the model. Anthropic also changed how Opus communicates, making it less likely to use jargon and more likely to lead with important information. Opus 5.5 is comparable to Mythos in biology and cybersecurity capabilities, so it carries the same safeguards as Fable, limiting use for discovering exploits in compiled programs or developing recognizable biological weapons.

The launch is Anthropic's first since CEO Dario Amodei embraced calls to pace the frontier, deliberately slowing capability progress to let alignment and risk prevention keep up. Whether that stance shapes future releases may hinge on whether the promised Sonnet 5.5 and Haiku 5.5 arrive with similar performance gains in the coming weeks.

FAQ
How much does Opus 5.5 cost?
Output tokens are charged at $20 per million tokens, compared to $25 for the previous model, with similar price drops on other metrics.
What safeguards apply to Opus 5.5?
Anthropic says Opus 5.5 is comparable to Mythos in biology and cybersecurity, so it carries the same safeguards as the Fable model, limiting exploit discovery and development of recognizable biological weapons.
When will Sonnet 5.5 and Haiku 5.5 be released?
According to Anthropic's announcement, they will be released 'in the coming weeks,' with similar performance improvements.

Get the latest Large Language Models news every morning

For example, today's edition would include:

  • Meta AI agent wins over NYT columnistTop Companies AI · 58m ago
  • Cyber vendor launches service to fight AI attacksTop Companies AI · 58m ago
  • Palo Alto Networks unveils AI cybersecurity service using Claude, GPTTop Companies AI · 58m ago

AI-summarized, only the topics you pick — one digest a day via Email, LINE, or Slack.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleClaude Opus 5.5 lands on Amazon Bedrock, Claude Platform on AWS