ModelRadar
/

Claude Opus 5.5

Anthropic🇺🇸 USFrontier· Sep 22, 2026

TextReasoningVisionAPI

Anthropic's flagship model with significantly improved price–performance ratio: it communicates more directly than its predecessor, matches Fable 5.1's intelligence at lower cost, and achieves higher token efficiency, making it suitable for extended reasoning and agentic workflows.

Summary of the model card

Context
1M tokens
Price / 1M tok
$2 in · $10 out
Modalities
text, image, document → text

2 variants

Standard1M$4/$20OR
Batch1M$2/$10OR

Sources

OpenRouterBlog
Deutsch

Das aktuelle Flaggschiff von Anthropic mit deutlich verbessertem Preis-Leistungs-Verhaeltnis. Es antwortet praeziser als sein Vorgaenger, bietet vergleichbare Intelligenz zu niedrigeren Kosten und eignet sich gut fuer ausgedehnte Reasoning- und Agenten-Workflows. Bei maximaler Denkhoechstgrenze kann es jedoch unzuverlaessig werden.

Provenance

Detected
Sep 27, 2026
First source
openrouter
Description
Summary of the model card
Editorially reviewed
—