Claude Opus 4.6 Fast
About this model
Claude Opus 4.6 Fast is a latency-optimized configuration of Anthropic's Claude Opus 4.6, the top tier of the Claude family above the Sonnet and Haiku lines. It runs the same underlying model through a "fast mode" path, keeping the headline capabilities — a one-million-token context window (in beta), up to 128K output tokens, adaptive thinking, and context compaction — while targeting lower latency for interactive and agentic use. Capabilities here span reasoning, vision, code-optimized work, function calling, and web search.
Compared with its same-family predecessor Claude Opus 4.5, Anthropic positions Opus 4.6 as a meaningful step in long-context performance and deeper reasoning; partners noted it "thinks longer" and handles careful exploration like debugging better than Opus 4.5. On Anthropic's own BigLaw Bench it reportedly scored 90.2%, the highest of any Claude model at release, and the ARC Prize Foundation reported 94.00% on ARC-AGI-1.
The standard, full-quality version is Claude Opus 4.6; this Fast variant prioritizes responsiveness over maximum deliberation, so very hard reasoning tasks may still favor the full model.
Within the speed-optimized lineage, Opus 4.6 Fast is the earliest entry, later followed by Claude Opus 4.7 Fast and Claude Opus 4.8 Fast. Anthropic's documentation recommends the newer Opus generations for the most complex, long-horizon agentic work.
This About section is AI-generated from public sources (Claude Opus 4.8), with no human editing. It may contain inaccuracies — verify critical details against the sources listed above.
Data sources: Venice API · HuggingFace · Wikipedia — enrichment updated 1d ago