Ai Model·claude-opus-4-5·Record·Profile

AI Model: Claude Opus 4.5 (anthropic)

Verdictpartial85%

1 check · 4/30/2026

1 → partial

Our claim

entire record

Entity Type: ai-model
Title: Claude Opus 4.5
Description: Claude Opus 4.5 was released November 24, 2025. First model to break 80% on SWE-bench Verified (80.9%). Featured a 67% price reduction from Opus 4 (\$5/\$25 vs \$15/\$75 per MTok). Scored 44% on Terminal-Bench Hard and 70 on Artificial Analysis Intelligence Index.
Tags: claude,claude-4,agentic-coding
Last Updated: November 2025
Related Entries: [object Object],[object Object]
Metadata: [object Object]

Source evidence

1 src · 1 check

www.anthropic.com/news/claude-opus-4-5 resource

partial85%Haiku 4.5 · 4/30/2026

NoteThe source confirms title (Claude Opus 4.5), developer (Anthropic), releaseDate (Nov 24, 2025), inputPrice ($5 per million tokens), outputPrice ($25 per million tokens), and contextWindow (200,000 tokens). However, the record claims 'safetyLevel: ASL-3', which is NOT mentioned anywhere in the source text. The source extensively discusses safety improvements and alignment but never uses the designation 'ASL-3' or any comparable safety level classification. The source states Opus 4.5 is 'the most robustly aligned model we have released to date' but does not assign it a formal safety level code. This is a partial confirmation because most fields are verified but one critical field (safetyLevel) is unverifiable from the source.

Case № claude-opus-4-5Filed 4/30/2026Confidence 85%