Claude Family

Claude is a family of large language models (LLMs) developed by Anthropic PBC since 2023. The name comes from Claude Shannon, father of information theory. The family is characterized by two distinctive features: (1) training based on Constitutional AI with RLAIF (Reinforcement Learning from AI Feedback), delivering safe and helpful model behavior with far fewer human annotators than classic RLHF; (2) exceptionally strong coding and long-form reasoning capabilities. The family spans three tiers: Haiku (fastest, cheap queries), Sonnet (performance/cost balance, flagship for coding), Opus (largest, deep reasoning). Current releases (2026): Claude Opus 4.5, 4.6, 4.7, 4.8; Claude Sonnet 4.5 and 5; Claude Haiku 4.5. Specialised variants: Claude Fable 5 and Claude Mythos 5 (long context up to 1M tokens, storytelling). Models are available via the Anthropic API, AWS Bedrock, Google Cloud Vertex AI, and the Claude.ai / Claude Desktop / Claude Code applications. Claude 3.7 Sonnet (2025) introduced extended thinking - inline chain-of-thought with visible reasoning steps. From the start, the family has been positioned for the enterprise segment (Slack, Notion, Zoom, GitHub Copilot use Claude as an option).