Claude Opus 4 moves to version Claude Opus 4 (claude-opus-4-20250514)

The OnAirToday AI Progress Tracker recorded a verified change to Anthropic's Claude Opus 4 (version Claude Opus 4 (claude-opus-4-20250514)): Version now Claude Opus 4 (claude-opus-4-20250514) (was claude-opus-4).

Claude Opus 4 is Anthropic’s May 2025 reasoning model for demanding coding, research, and tool-using workflows, with a 200,000-token context window. Later Opus generations have shipped; the specifications and benchmark results here remain specific to Opus 4.

What changed in this review:

• Added the documented 200,000-token context window.
• Added launch API pricing: $15 input and $75 output per million tokens, excluding caching and batch discounts.
• Replaced GPQA Diamond and SWE-bench Verified figures with Anthropic’s published launch results.
• Added published AIME 2025, MMMU, Terminal-Bench, and domain-specific Tau-bench results.
• Added documented agent limitations, safety-evaluation findings, and ASL-3 deployment protections.
• Claude Opus 4.1 shipped after Opus 4; this profile remains specific to the original model.

Teams building on Claude Opus 4 should confirm limits, pricing and benchmark coverage against the vendor's own documentation before relying on these figures.

The full profile — capabilities, benchmark citations, known flaws and risk guidance — is on the tracker at /ai-models/claude-opus-4.

References:

• https://www.anthropic.com/news/claude-4
• https://www.anthropic.com/claude-4-system-card
• https://www.anthropic.com/news/activating-asl3-protections
• https://www.anthropic.com/news/claude-opus-4-1