Skip to content

Claude

Models

13 Claude pages in Models.

Claude 3 ModelsThe Claude 3 family launched on Mar 4, 2024 with Claude 3 Opus, Sonnet and Haiku, followed by Claude 3.5 Sonnet (Jun 21, 2024), its upgrade and Claude 3.5 Haiku (Oct 22, 2024), and Claude 3.7 Sonnet (Feb 24, 2025), the first Claude with extended thinking. All of them are retired on the Claude API; the last, Claude 3 Haiku, on Apr 20, 2026.
Claude 4 ModelsThe Claude 4 family ran from Claude Opus 4 and Sonnet 4 (May 22, 2025) to Claude Opus 4.8 (May 28, 2026). As of Nov 6, 2026, Opus 4.5 to 4.8, Sonnet 4.6 and Haiku 4.5 are still served on the Claude API, Sonnet 4.5 retires on Nov 30, 2026, and Opus 4, Sonnet 4 and Opus 4.1 are retired.
Claude BenchmarksAs of Nov 6, 2026, Claude Opus 5.5 leads Terminal-Bench 4.0 and WebDev Arena, ranks third on Arena's text board behind two Gemini 4 models, second on the Artificial Analysis index, and third on ARC-AGI-2 behind two OpenAI models. Anthropic's own launch tables run a few points higher than the public boards.
Claude Context WindowEvery current Claude model (Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5) has a 1M-token context window on the API, about 555,000 words, and can write up to 128K tokens in one response. In the Claude apps the window is up to 1M on paid plans depending on the model, and long chats are summarized automatically near the limit.
Claude Fable 5Claude Fable 5 launched on Jun 9, 2026 as Anthropic's first Mythos-class model for general use, was suspended for all users on Jun 12 when US export controls applied to it, and returned on Jul 1, 2026. Claude Fable 5.1 replaced it as the current model in Sep 2026; Fable 5 remains on the API as a legacy model.
Claude Fable 5.1Claude Fable 5.1 is Anthropic's most capable widely released model, released on Sep 1, 2026 and priced at $10 input and $50 output per million tokens on the Claude API. It is included on Claude Max, Ultra, Team premium seats and Enterprise, and paid with usage credits on Pro.
Claude Haiku 5.5Claude Haiku 5.5 is Anthropic's fastest and cheapest current model, released on Oct 7, 2026 at $0.10 input and $0.50 output per million tokens for prompts up to 100,000 tokens. It suits classification, extraction, routing and subagents, and is on every Claude plan including Free.
Claude Model HistoryAnthropic released the first Claude model on Mar 14, 2023 and has shipped more than 30 since, through Claude 2, the Claude 3, 3.5 and 3.7 models, the Claude 4 family and the Claude 5 family that is current in Nov 2026. This page lists every release with its date, what it added and when it was retired.
Claude ModelsAnthropic sells four current Claude models: Claude Fable 5.1, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5, all with a 1M-token context window. Anthropic's docs suggest starting with Opus 5.5 and moving up to Fable 5.1 only when a task needs it; Claude Mythos 5.1 is restricted to verified organizations.
Claude MythosClaude Mythos is Anthropic's name for its most capable models with some safeguards lifted, offered only to vetted organizations for defensive cybersecurity and life sciences work. It began as Claude Mythos Preview for Project Glasswing on Apr 7, 2026; the current version, Claude Mythos 5.1, is the same model as Claude Fable 5.1 and is limited to US organizations in Anthropic's verification programs.
Claude Opus 5.5Claude Opus 5.5 is Anthropic's recommended starting model, released on Sep 22, 2026 at $4 input and $20 output per million tokens on the Claude API, with a 1M-token context window. It is on every paid Claude plan and is the default model in Claude Code.
Claude Sonnet 5.5Claude Sonnet 5.5 is Anthropic's mid-priced model, released on Sep 28, 2026 at $2 input and $10 output per million tokens with a 1M-token context window. It is on every Claude plan including Free, and costs half as much as Claude Opus 5.5.
Extended and Adaptive ThinkingThinking lets Claude reason before it answers. On current Claude models it is adaptive (the model decides how much to think) and is steered with an effort level from Low to Max: in the apps through the model menu next to the send button, on the API with output_config.effort. Thinking tokens are billed as output.