Claude Sonnet 4.5 moves to version claude-sonnet-4-5-20250929

The OnAirToday AI Progress Tracker recorded a verified change to Anthropic's Claude Sonnet 4.5 (version claude-sonnet-4-5-20250929): Version now claude-sonnet-4-5-20250929 (was claude-sonnet-4-5).

Claude Sonnet 4.5 is Anthropic's hybrid-reasoning model for coding, tool use and computer-based agents, with text and image inputs. Later releases in the same model line include Claude Sonnet 4.6.

What changed in this review:

• Claude Sonnet 4.6 shipped on February 17, 2026; this entry continues to describe Sonnet 4.5.
• Clarified the 200K standard context window and optional 1M-token extended-context beta.
• Anthropic's launch results report GPQA Diamond at 83.4% and SWE-bench Verified at 77.2%.
• Terminal-Bench is reported at 50.0% on the original benchmark, not Terminal-Bench 2.0.
• Added AIME 2025 performance without tools.
• Added documented limitations and Claude Code cyberespionage misuse; the incident's underlying model version was not disclosed.

Teams building on Claude Sonnet 4.5 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-4-5.

References:

• https://www.anthropic.com/news/claude-sonnet-4-5
• https://platform.claude.com/docs/en/about-claude/pricing
• https://www.anthropic.com/news/claude-sonnet-4-6
• https://www.anthropic.com/news/disrupting-AI-espionage