Claude models list: every Anthropic model by date

Anthropic’s Claude models, as of October 23, 2026, come in four generally available versions: Claude Fable 5.1, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5. A fifth, Claude Mythos 5.1, is the same model as Fable 5.1 with different safeguards and is open only to organisations in Anthropic’s trusted access programs. Below is the current line-up with its model IDs and limits, followed by every Claude model Anthropic has released since Claude 4 in May 2025, each dated from Anthropic’s own announcement or Claude Platform release notes.

The current Claude line-up

Anthropic’s models overview lists four current models. All four take text and image input, return text, support tool use and share the same context and output limits:

Source: Anthropic models overview and model deprecations pages, checked October 23, 2026.

Anthropic describes the four by job. Fable 5.1 is “for demanding reasoning and long-horizon agentic work”, Opus 5.5 “for long-running agentic coding and knowledge work”, Sonnet 5.5 is “the best combination of speed and intelligence”, and Haiku 5.5 is for “high-volume, latency-sensitive tasks such as classification, extraction, and routing”. The docs rank their latency from slower (Fable 5.1) through moderate (Opus 5.5) and fast (Sonnet 5.5) to fastest (Haiku 5.5).

The docs’ own starting advice is to use Opus 5.5 for most workloads and to move up to Fable 5.1 when evaluations on Opus 5.5 at higher effort still fall short. Thinking is adaptive on all four and always on for Fable 5.1 and Opus 5.5. Default effort is high on Fable 5.1 and Sonnet 5.5 and medium on Opus 5.5 and Haiku 5.5. Haiku 5.5 is the first Haiku-class model with an adjustable effort setting, according to Anthropic’s launch post.

The same IDs work on the Claude API, Claude Platform on AWS, Microsoft Foundry and both Google Cloud routes; Amazon Bedrock prefixes them with anthropic., as in anthropic.claude-opus-5-5.

Four lines: Haiku, Sonnet, Opus and Fable

Claude model names carry a line and a version number. Haiku, Sonnet and Opus are the long-running small, mid-size and large lines. Fable and Mythos arrived in 2026 above Opus.

Mythos started as a restricted model. On April 7, 2026, Anthropic announced Claude Mythos Preview as a gated research preview for defensive cybersecurity work under Project Glasswing, with access by invitation only. Anthropic said Mythos Preview could find and exploit software vulnerabilities better than all but the most skilled humans, and that this was why it kept the release limited.

Fable is the general-availability version of that class. On June 9, 2026, Anthropic launched Claude Fable 5, calling it “a Mythos-class model that we’ve made safe for general use”, alongside Claude Mythos 5 for Project Glasswing participants. With Fable 5.1 and Mythos 5.1, Anthropic states it outright: “Claude Fable 5.1 and Claude Mythos 5.1 are the same model, but with different levels of safeguards.” Fable 5.1 is generally available; Mythos 5.1 goes through trusted access programs, with safeguards designed for work in cybersecurity and the life sciences.

So a newer version number does not always mean a stronger model across lines. Anthropic says Opus 5.5 “performs at the level of Claude Fable 5.1 on most work”, while still placing Fable 5.1 above it for the hardest reasoning and agentic tasks. Within a line, a higher number replaces the lower one. For programs that need to group models, the Models API now returns a line field for each model, and the docs advise reading it rather than inferring the line from the model ID.

Every Claude model by date, since Claude 4

Dates are Anthropic’s launch dates from its announcements and Claude Platform release notes. Status is from Anthropic’s model deprecations page for Anthropic-operated platforms.

Two patterns stand out. The gap between versions has shortened: Opus went from 4.5 to 5.5 in ten months, through 4.6, 4.7, 4.8 and 5. And older models stay available longer than the release pace suggests: on October 23, 2026, Anthropic still lists Opus 4.5, 4.6, 4.7 and 4.8 as active, next to Opus 5 and Opus 5.5.

Older Claude models, all retired

Anthropic’s deprecations page records the end of every model before Claude 4. Requests to a retired model fail.

The next scheduled retirement is Claude Sonnet 4.5 on November 30, 2026. Anthropic notified developers on September 30 and recommends Sonnet 5.5 as the replacement.

How Claude models are retired

Anthropic uses four lifecycle terms. An active model is fully supported. A legacy model no longer receives updates and might be deprecated later. A deprecated model still works but is no longer recommended, and has a replacement and a retirement date. A retired model is no longer available.

Anthropic says it gives at least 60 days’ notice before retiring a publicly released model, and every current model’s entry carries a “not sooner than” date, one year after its launch. Those dates apply to Anthropic-operated platforms; Amazon Bedrock and Google-managed Claude on Google Cloud set their own retirement schedules, so a model’s status there can differ. Anthropic has also committed to preserving the weights of retired models.

New Claude releases are logged as they happen in our AI model updates tracker, scores that Anthropic and benchmark maintainers publish are on the LLM leaderboard, and our research reviews of Claude Opus 5.5 and Claude Sonnet 5.5 collect independent evaluations. What labs disclose about each model is covered in our model cards explainer.