Claude Model Release Timeline: Every Claude Version and Its Release Date (2023–2026)
Live timeline · last updated · 1 entry in the update log
A single dated list of Claude releases, built only from Anthropic’s news pages and the Claude Platform release notes. It covers the Opus, Sonnet and Haiku lines and the newer Mythos and Fable models, including the June 2026 suspension.
As of Oct 21, 2026 the newest Claude models are Claude Haiku 5.5 (Oct 7), Claude Sonnet 5.5 (Sep 28), Claude Opus 5.5 (Sep 22) and Claude Fable 5.1 (Sep 1), with Claude Mythos 5.1 restricted to verified cyber and life-science organizations; the first Claude shipped on Mar 14, 2023.
Read on 2026-10-21: the anthropic.com news page for each model listed, Anthropic’s Project Glasswing page, its June 12 statement on the suspension, the Fable 5.1 and Mythos 5.1 page, and the Claude Platform release notes. Dates are the publication dates shown on those pages. We list models, not every snapshot or feature release, and we ran no benchmarks.
Timeline
-
Anthropic introduces Claude, available through a chat interface and an API in its developer console, in two versions: Claude and the lighter, faster and cheaper Claude Instant. This is the first public Claude model.
-
Claude 2 is released with better performance and longer responses, available through the API and on a new public beta website, claude.ai.
-
Claude 2.1 brings a 200K-token context window, lower hallucination rates, system prompts and a beta of tool use.
-
The Claude 3 family is announced in three sizes, Haiku, Sonnet and Opus. Opus and Sonnet are available in the API the same day; the post says Haiku will follow soon.
-
Claude 3.5 Sonnet is released as the first model of the Claude 3.5 family, alongside a new Artifacts feature on claude.ai. The news page is dated June 21, 2024.
-
Anthropic ships an upgraded Claude 3.5 Sonnet, announces a new Claude 3.5 Haiku for release later that month, and puts “computer use”, where Claude operates a desktop through screenshots, mouse and keyboard, into public beta.
Anthropic: Computer use, Claude 3.5 Sonnet, and Claude 3.5 Haiku
-
Claude 3.7 Sonnet arrives as a hybrid model that can answer quickly or show extended step-by-step thinking, and Claude Code is introduced as a research preview.
-
Claude 4 launches as two models, Claude Opus 4 and Claude Sonnet 4, both with extended thinking.
-
Claude Opus 4.1 ships at the same price as Opus 4, with Anthropic reporting 74.5% on SWE-bench Verified.
-
Claude Sonnet 4.5 is released, described in the release notes as Anthropic’s best model for complex agents and coding at that time.
-
Claude Haiku 4.5 is released as the fastest Haiku model, aimed at real-time and high-volume work.
-
Claude Opus 4.5 is released, completing the 4.5 generation of Opus, Sonnet and Haiku.
-
Claude Opus 4.6 launches for complex agentic and long-horizon work. The release notes say it recommends adaptive thinking and that manual thinking budgets are deprecated.
-
Claude Sonnet 4.6 launches as the balanced everyday model, with extended thinking and a 1M-token context window in beta according to the release notes.
-
Anthropic announces Claude Mythos Preview, a model it calls strikingly capable at computer security, and does not release it generally. It goes instead to Project Glasswing, a defensive effort with Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks.
-
Claude Opus 4.7 launches at the same $5 / $25 per million tokens as Opus 4.6, with an updated tokenizer.
-
Claude Opus 4.8 launches with a 1M-token context window by default and 128k max output tokens.
-
Claude Fable 5 and Claude Mythos 5 launch. Anthropic says they are the same underlying model: Fable 5 is generally available with safeguards that hand some cyber, biology and distillation requests to Claude Opus 4.8, while Mythos 5 lifts some of those safeguards and goes only to Project Glasswing partners. Both cost $10 / $50 per million tokens.
-
Anthropic suspends access to Fable 5 and Mythos 5 for all customers. Its statement says a US government export control directive, received at 5:21pm ET, barred any foreign national from accessing the two models, and that complying meant disabling them for everyone. Other Claude models were not affected.
Anthropic: Statement on the directive to suspend Fable 5 access
-
Claude Sonnet 5 launches at introductory pricing of $2 / $10 per million tokens, made the standard price on August 10. The same day Anthropic publishes “Redeploying Claude Fable 5”.
-
The release notes record that access to Claude Fable 5 and Claude Mythos 5 has been restored.
-
Claude Opus 5 launches at $5 / $25 per million tokens with a 1M-token context window and 128k max output tokens.
-
Claude Fable 5.1 and Claude Mythos 5.1 launch, again one model with two safeguard levels. Prices stay at $10 / $50, but cache reads drop to $0.25 per million tokens, which Anthropic estimates cuts typical costs by about 25%.
-
Claude Opus 5.5 launches at $4 / $20 per million tokens, below the $5 / $25 of Opus 5.
-
Claude Sonnet 5.5 launches on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud and Microsoft Foundry.
-
Anthropic expands its Cyber Verification Program into three access tiers, each including Claude Opus 5.5, Claude Sonnet 5.5 and Claude Mythos 5.1. Existing Project Glasswing members move to the Specialized Access tier.
-
Claude Haiku 5.5 launches for high-volume and latency-sensitive work, with a 1M-token context window and 128k max output tokens.
How to read this list
Each entry is the day Anthropic published the announcement or release-notes entry for a model. When a model reached the API and the apps on different days, we use the date on Anthropic’s own announcement and say so in the text. Prices are list prices per million input and output tokens as Anthropic stated them at launch; they change, so check the current pricing page before you budget.
We list model generations and named variants. We do not list every dated snapshot, regional rollout or feature such as a new tool, unless it changed who could use a model, as with the June 2026 suspension and the October 2026 Cyber Verification Program tiers.
The naming, in one paragraph
From Claude 3 in March 2024 onward, Anthropic sold three sizes per generation: Haiku (small and fast), Sonnet (balanced) and Opus (largest). Version numbers moved in half and decimal steps, 3, 3.5, 3.7, 4, 4.1, 4.5 and so on, and the three sizes did not always ship together: Opus 4.5 followed Sonnet 4.5 by eight weeks. In 2026 Anthropic added a tier above Opus. Its own footnote calls Mythos-class models “a tier of Claude models that sit above our Opus class in capability”. Fable and Mythos are the same model with different safeguards, and the names come from the Latin fabula and the Greek mythos.
2026 in detail: the Mythos tier and the suspension
The year’s turning point is April 7, 2026, when Anthropic announced Claude Mythos Preview and held it back from general release. The Glasswing page says the model showed that AI had reached a level of coding capability where it could surpass all but the most skilled humans at finding and exploiting software vulnerabilities, so access went to a defensive coalition instead of to all customers.
On June 9 Anthropic shipped the first general Mythos-class model, Fable 5, with classifiers that route some cybersecurity, biology and chemistry, and distillation-related requests to Claude Opus 4.8. Anthropic said the fallback happened in under 5% of sessions. The unrestricted twin, Mythos 5, went to Glasswing partners. Three days later, on June 12, Anthropic disabled both models for every customer after a US export control directive barred foreign nationals from using them. The release notes record access restored on July 1.
From there the cadence sped up. Sonnet 5 came on June 30, Opus 5 on July 24, Fable 5.1 and Mythos 5.1 on September 1, Opus 5.5 on September 22, Sonnet 5.5 on September 28 and Haiku 5.5 on October 7. That is seven model launches in about fourteen weeks. On October 6 the Cyber Verification Program became the main way to get Mythos 5.1, with three access tiers.
Which Claude model is the newest, by line
- Opus: Claude Opus 5.5, launched September 22, 2026, at $4 / $20 per million tokens.
- Sonnet: Claude Sonnet 5.5, launched September 28, 2026.
- Haiku: Claude Haiku 5.5, launched October 7, 2026, with a 1M-token context window.
- Fable (general Mythos-class): Claude Fable 5.1, launched September 1, 2026, at $10 / $50.
- Mythos (restricted): Claude Mythos 5.1, also September 1, 2026, available only through Anthropic’s verification programs.
Gaps between releases
Measured from announcement to announcement, the time between flagship Opus releases shrank sharply. Opus 3 (March 2024) to Opus 4 (May 2025) took about fourteen months. Opus 4 to Opus 4.1 took about ten weeks, Opus 4.5 to 4.6 about ten weeks, and Opus 5 to Opus 5.5 about eight and a half weeks. Sonnet followed a similar curve. A shorter gap does not mean a smaller change: the release notes call Opus 5 “a step-change improvement over Claude Opus 4.8”.
Price moved in both directions. Opus 4.6, 4.7 and Opus 5 stayed at $5 / $25, and Opus 5.5 dropped to $4 / $20. The Mythos-class models launched at $10 / $50, which Anthropic described as less than half the price of Claude Mythos Preview, and Fable 5.1 cut cache-read prices instead of headline prices.
What we did not include, and why
- Claude Instant versions after the first one, and dated snapshots of a model. They are versions of a listed model, not new generations.
- Any date reported only by press or social media. Every node above points to an Anthropic page.
- Benchmark claims. Anthropic’s pages contain many, but we did not reproduce them, so we cite none here.
- Future models. Anthropic has said more capable models are coming; we add a model only when Anthropic publishes it.
Related timelines
The same format covers the other frontier vendors: the OpenAI GPT model timeline, the Gemini model timeline and the Grok model timeline. For a narrower story told the same way, see Terence Tao vs OpenAI. The Front blog index lists every post.
The same format also covers Moonshot Kimi, Mistral, DeepSeek and Alibaba Qwen.
Keep reading on Front
Every post, newest first, is on the Front blog index.
- OpenAI GPT model release timeline: every GPT and o-series model from ChatGPT to GPT-6.
- Gemini model release timeline: every Gemini version from 1.0 to Gemini 4 Argon.
- Grok model release timeline: every Grok version, with Elon Musk’s posts quoted and linked.
- Meta Muse model timeline: Muse Spark to Muse Wonder, with prices.
- AI model releases in 2024: the 2024 nodes from the vendor timelines, in date order.
- AI model releases in 2025: the 2025 nodes from the vendor timelines, in date order.
- AI model releases in 2026: the 2026 nodes from the vendor timelines, in date order.
- Kimi model timeline: Moonshot’s Kimi line from K1.5 to K3.1 Pro, with licenses and context.
FAQ
What is the latest Claude model?
As of October 21, 2026: Claude Haiku 5.5 (October 7), Claude Sonnet 5.5 (September 28), Claude Opus 5.5 (September 22) and Claude Fable 5.1 (September 1). Claude Mythos 5.1 is available only to verified organizations.
When was Claude Mythos released?
Claude Mythos Preview was announced on April 7, 2026 for Project Glasswing partners only. Claude Mythos 5 followed on June 9, 2026 and Claude Mythos 5.1 on September 1, 2026, both restricted.
What is the difference between Claude Fable and Claude Mythos?
Anthropic says they are the same underlying model. Fable is generally available with stricter safeguards; Mythos lifts some safeguards for vetted cybersecurity and life-science users.
When was the first Claude released?
March 14, 2023, as Claude and Claude Instant.
Update log
- — First published. Checked Anthropic’s news pages and the Claude Platform release notes through Oct 21, 2026 (UTC+8).
Sources
- Anthropic: Introducing Claude
- Anthropic: Introducing the next generation of Claude
- Anthropic: Introducing Claude 4
- Anthropic: Project Glasswing
- Anthropic: Claude Fable 5 and Claude Mythos 5
- Anthropic: Statement on the directive to suspend Fable 5 access
- Anthropic: Claude Fable 5.1 and Mythos 5.1
- Anthropic: Introducing Claude Opus 5.5
- Anthropic: Introducing Claude Haiku 5.5
- Anthropic: Expanding the Cyber Verification Program
- Claude Platform release notes