All Claude models, always current

Anthropic's newest top-end model. 1M-token context, adaptive thinking, the strongest tool-use accuracy in the lineup.

Best price-to-performance Claude. 1M-token window, full tool-use, extended thinking — for daily production traffic.

Prior flagship. Still excellent for long-context agentic work if you don't need 4.7's adaptive thinking.

200k-token Opus generation. Still widely used in tools that haven't migrated to the 4.6+ 1M-context family.

Fastest Claude. For chatbots, classification, and high-volume routing where TTFT matters.

Prior Sonnet generation with 1M context. Strong general-purpose model.

Earlier Opus 4 series. Available for cost or reproducibility reasons.

Original Claude Opus 4 launch. Pinned for legacy compatibility.

Original Sonnet 4 with the 1M context window. Stable choice if 4.6+ isn't required.