AIster · The agentic age · Launches

Anthropic releases Claude Opus 4 and Claude Sonnet 4

Anthropic's first model released under its stricter ASL-3 protections

Key facts

Date

22 May 2025

Era

The agentic age

Interest

Launches

Threads

Model wars, Safety & governance, Labs & people

Who

Anthropic

API price

Opus 4 $15/$75, Sonnet 4 $3/$15 per million tokens

Safety

ASL-3 activated for Opus 4 as a precaution

On 22 May 2025 Anthropic released Claude Opus 4 and Claude Sonnet 4, aimed at coding and long-running agent work. It activated its ASL-3 (AI Safety Level 3) deployment and security standards for Opus 4 as a precaution, focused on chemical, biological, radiological and nuclear weapons risks, while saying it had not determined that the model definitely required them.

Sources

1. Anthropic, Introducing Claude 4, 2025-05-22
2. Anthropic, Activating AI Safety Level 3 protections, 2025-05-22