Claude Opus 5.5 is Anthropic’s most capable Opus release, built for agent systems that run for hours, large-codebase engineering, and analysis that spans many documents. It reads text and images, uses tools, and exposes configurable effort and extended thinking, so a team can trade latency against depth on a per-request basis. Its 1M-token context window and 128K-token maximum output suit codebase-wide refactors, research synthesis, and long transcript work.

Claude Opus 5 is Anthropic's high-capability model for complex agentic coding, deep reasoning, long-running tool use, and enterprise knowledge work. It combines a 1M-token context window with vision, configurable effort levels, and reasoning enabled by default, allowing teams to balance latency and cost against capability. It is particularly suited to large codebases, multi-step research and analysis, document-heavy workflows, and agent systems that need sustained, reliable execution.

Claude Fable 5 is Anthropic's new top-capability generally available Claude model. It is built for long-running coding, agentic execution, multimodal reasoning, research, and high-stakes professional workflows, with stronger long-horizon performance than earlier Opus models, state-of-the-art vision, and a 1M-token context window. Safety classifiers can route some sensitive requests to Claude Opus 4.8 instead.

Claude Opus 4.8 is Anthropic's highest-capability Claude model. It is built for demanding coding, agent orchestration, multimodal reasoning, and professional workflows that need strong instruction following, adaptive and extended thinking, high-resolution vision, and a 1M-token context window.

Claude Opus 4.7 is Anthropic's highest-capability generally available Claude model. It is designed for demanding coding, agent orchestration, multimodal reasoning, and high-stakes professional workflows, with stronger instruction following, better high-resolution vision, adaptive thinking, and a 1M-token context window.

Claude Sonnet 4.6 is Anthropic's most capable Sonnet model, built for daily production use across coding, agent workflows, long-context reasoning, computer use, and professional knowledge work. It supports adaptive and extended thinking, strong instruction following, high-volume automation, and a 1M-token context window in beta.

Claude Haiku 4.5 is Anthropic's fastest and most cost-efficient Claude model. It is built for latency-sensitive applications, high-volume agents, sub-agent orchestration, coding assistance, and budget-conscious deployments that still need strong reasoning and multimodal understanding.

Claude Fable 5.1 is the next release in Anthropic’s Fable line, aimed at long-horizon coding, agentic execution, and multimodal research. It reads text and images, calls tools, and supports extended thinking, with a 1M-token context window and a 128K-token maximum output for work that has to hold an entire project in view. It fits document-heavy analysis, sustained agent runs, and engineering tasks that span many files.