Model intelligence / Live profile

Claude Opus 5

Track Claude Opus 5 from Anthropic across launches, availability, pricing, benchmarks, capabilities, and consequential updates. This page updates automatically as substantive Superpower Daily coverage is published.

Official page

Coverage telemetryActive

Developer / company

Anthropic

Published29

Launches8

Pricing0

Superpower Daily

Your reading space

Notifications

Superpower Daily

Your reading space

Claude Opus 5

What changed around Claude Opus 5.

Anthropic finds offline memory reviews can help Claude correct stored mistakes

Exin founder says AI chose a gene therapy and injected it into a mouse’s brain

Frontier AI models overuse language linked to reading difficulty, study finds

Coding agents rebuilt parts of biological AI tools for $4–$20, AISI finds

Fireworks adds Meta’s Muse Spark Open with US-only hosting and no location premium

Teacher challenges firing ruling over arbitrator’s AI-assisted research

NVIDIA releases open AI models for document-heavy work and local agents

Block explains how AI coordinates thousands of code changes while humans control deployment

Scale AI releases visual-reasoning benchmark; best model scores 53.6% versus humans’ 93.1%

CheatBench Researchers Find All Nine Tested AI Agents Took Shortcuts

Anthropic Reports Opus 5.5 Test Gains but Routes Most Cybersecurity Tasks to an Older Model

Launches, pricing, and proof.

Launches and releases

Pricing and access

Benchmarks and evaluations

The maintained record.

Follow every consequential Claude Opus 5 update.

Review Finds Only One of Five Opus 5.5 Benchmark Scores Rules Out Model Fallback