Claude Sonnet 4.5 shipped on September 29, 2025; this profile remains specific to Claude Sonnet 4

The OnAirToday AI Progress Tracker recorded a verified change to Anthropic's Claude Sonnet 4 (version claude-sonnet-4-20250514): Claude Sonnet 4.5 shipped on September 29, 2025; this profile remains specific to Claude Sonnet 4.

Claude Sonnet 4 is Anthropic’s May 2025 hybrid-reasoning model, supporting coding, multimodal analysis and tool-assisted workflows. Claude Sonnet 4.5 subsequently shipped on September 29, 2025; the specifications and benchmark results here remain specific to Sonnet 4.

What changed in this review:

• The published mathematics result is identified as AIME 2025.
• The published τ-bench result is identified as the retail evaluation, not an aggregate across domains.
• Added Anthropic’s documented Claude Code extortion incident, without attributing it to a confirmed Sonnet checkpoint.
• Capability score 64 → 68
• Flaws and limitations updated
• Risks and threats updated

Teams building on Claude Sonnet 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-sonnet-4.

References:

• https://www.anthropic.com/news/claude-4
• https://www.anthropic.com/news/claude-sonnet-4-5
• https://docs.anthropic.com/en/docs/build-with-claude/context-windows
• https://www.anthropic.com/news/detecting-and-countering-misuse-of-ai-august-2025