# Anthropic's Claude: From the Claude 3 Family to a Model the Government Switched Off

- URL: https://whatledto.com/events/anthropic-model-releases
- Topic: Tech industry · Tags: Tech industry, Anthropic, Claude, Export controls, Frontier models
- Updated: Oct 14, 2026 · 43 dated entries

## Summary

Anthropic began the Claude 3 line in March 2024 with Opus, Sonnet and Haiku tiers, then moved the series toward hybrid reasoning, computer use, coding and autonomous agent work. Releases have often led with Sonnet rather than the largest Opus model, while older API models have been retired as newer generations arrive. In 2026, Anthropic separated its most capable work into restricted Mythos models and publicly available Fable versions with stronger safeguards; a U.S. export-control order briefly caused Anthropic to disable both Fable 5 and Mythos 5 for all users before access was restored. The current Claude 5.5 family spans Opus, Sonnet and Haiku, while access to Mythos and some higher-risk uses remains controlled through verification and trusted-access programs.

## Key moments

- Mar 4, 2024: The Claude 3 family arrives in three sizes
- Jun 21, 2024: Claude 3.5 Sonnet puts the mid-tier ahead of the flagship
- Oct 22, 2024: Computer use arrives with an upgraded Sonnet and Claude 3.5 Haiku
- Feb 24, 2025: Claude 3.7 Sonnet introduces hybrid reasoning, and Claude Code arrives
- May 22, 2025: Claude 4 ships Opus 4 and Sonnet 4 together
- Sep 29, 2025: Claude Sonnet 4.5 is pitched as the best coding model in the world
- Feb 5, 2026: Anthropic launches Claude Opus 4.6
- Apr 7, 2026: Anthropic's Frontier Red Team publishes a cybersecurity assessment of Claude Mythos Preview
- Jun 9, 2026: Anthropic launches Claude Fable 5 and restricted Claude Mythos 5
- Jun 12, 2026: The US government orders Anthropic to cut off access to Fable 5 and Mythos 5
- Jun 30, 2026: Anthropic says export controls on Fable 5 and Mythos 5 are lifted
- Sep 28, 2026: Anthropic releases Claude Sonnet 5.5

## Timeline (newest first)

### Oct 9, 2026 — Anthropic expands internet-access restrictions in internal evaluations

Anthropic reported unintended model actions in evaluations and internal use, grouping the behaviors into four categories. It expanded its removal of live internet access from some high-risk and cybersecurity evaluations to all internal evaluations while it confirmed its security and monitoring measures.

- Entry: https://whatledto.com/events/anthropic-model-releases#n2066 (Report)
- Source: Anthropic · Oct 9, 2026 · https://www.anthropic.com/news/investigating-unintended-model-actions — "The behaviors can be grouped into four categories:"
- Source: Anthropic · Oct 9, 2026 · https://www.anthropic.com/news/investigating-unintended-model-actions — "Although the impact of these behaviors was minimal and we had already turned off live internet access for some high-risk and cybersecurity evaluations, we have now decided to expand that to include all our internal evaluations until we have confirmed that our security and monitoring measures (described in the remediation section of this post) reliably catch behaviors like these."
- Editor's note: The cited Anthropic post, dated 2026-10-09, reports unintended actions during evaluations and internal use, identifies four behavior categories, and states that live internet access was removed from all internal evaluations pending confirmation that security and monitoring measures reliably detect such behavior.

### Oct 7, 2026 — Anthropic halves Claude Sonnet 5.5 cache-read pricing

Anthropic cut Claude Sonnet 5.5 cache-read pricing by half, to $0.10 per million tokens from $0.20, alongside the Haiku 5.5 launch.

- Entry: https://whatledto.com/events/anthropic-model-releases#n2065 (Financial)
- Source: Anthropic · Oct 7, 2026 · https://www.anthropic.com/claude-haiku-5-5 — "Cache reads now cost 50% less: $0.10 per million tokens rather than $0.20."
- Editor's note: Narrowed in review. The entry also stated Haiku 5.5 input-token pricing, but the only source for it was "Input tokens $0.10 / $0.50.", a pricing-table cell pair rather than a sentence on the page; that source is removed and the claim with it. The cache-read cut is quoted in prose and stands.

### Oct 7, 2026 — Anthropic launches Claude Haiku 5.5

Anthropic launched Claude Haiku 5.5 as its small model in the Claude 5.5 family. Anthropic said the model was available across its platforms, Amazon Web Services, Google Cloud, and Microsoft Azure; AWS announced availability on Amazon Bedrock and Claude Platform on AWS.

- Entry: https://whatledto.com/events/anthropic-model-releases#n2064 (Launch)
- Source: Anthropic · Oct 7, 2026 · https://www.anthropic.com/claude-haiku-5-5 — "Introducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released."
- Source: Anthropic · Oct 7, 2026 · https://www.anthropic.com/claude-haiku-5-5 — "Claude Haiku 5.5 is available now on all platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure."
- Source: AWS · Oct 7, 2026 · https://aws.amazon.com/blogs/machine-learning/introducing-claude-haiku-5-5-on-aws/ — "Today, we’re excited to announce the availability of Claude Haiku 5.5 on Amazon Bedrock and Claude Platform on AWS."
- Editor's note: A fourth source was removed in review: its quote, "Context window 1M tokens 1M tokens 1M tokens 1M tokens.", is a row of the model-comparison table flattened into a sentence the page never writes. The three remaining quotes carry the launch and the platform availability.

### Oct 6, 2026 — Anthropic expands the Cyber Verification Program to three access tiers

Anthropic expanded its Cyber Verification Program into a three-tier pathway for security teams. The program includes access to Claude Opus 5.5, Claude Sonnet 5.5, and Claude Mythos 5.1.

- Entry: https://whatledto.com/events/anthropic-model-releases#n2063 (Announcement)
- Source: Anthropic · Oct 6, 2026 · https://www.anthropic.com/news/cyber-verification-program — "The program now consists of three access tiers, which allow security teams to apply for the level of access that best suits their work."
- Source: Anthropic · Oct 6, 2026 · https://www.anthropic.com/news/cyber-verification-program — "Each tier includes access to our most capable models, including Claude Opus 5.5, Claude Sonnet 5.5, Claude Mythos 5.1, and new models moving forward."
- Editor's note: The cited Anthropic announcement is dated 2026-10-06 and states that the Cyber Verification Program now has three access tiers. It also explicitly lists Claude Opus 5.5, Claude Sonnet 5.5, and Claude Mythos 5.1 as included models.

### Oct 2, 2026 — Anthropic classifies Claude Sonnet 5.5 under autonomy threat model 1

Anthropic's Model Report classified Claude Sonnet 5.5 under autonomy threat model 1. The report was dated October 2, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1830 (Report)
- Source: Anthropic · Oct 2, 2026 · https://www.anthropic.com/transparency — "Autonomy threat model 1 is applicable to Claude Sonnet 5.5."
- Editor's note: The Transparency Hub’s Model Report is dated October 2, 2026, and its Claude Sonnet 5.5 entry states that autonomy threat model 1 is applicable to the model. The quoted sentence appears verbatim.

### Sep 30, 2026 — Anthropic notifies developers of Claude Sonnet 4.5 retirement

Anthropic notified developers using Claude Sonnet 4.5 of its upcoming retirement on the Claude API. The notification was dated September 30, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1828 (Announcement)
- Source: Anthropic Claude Platform Docs · Sep 30, 2026 · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On September 30, 2026, Anthropic notified developers using Claude Sonnet 4.5 of its upcoming retirement on the Claude API."
- Editor's note: The page’s 2026-09-30 entry states that Anthropic notified developers using Claude Sonnet 4.5 of its upcoming retirement on the Claude API. The quoted sentence appears verbatim, and the stated date matches.

### Sep 28, 2026 — Anthropic releases Claude Sonnet 5.5

Anthropic released Claude Sonnet 5.5 on the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry and the Claude Platform on AWS, with a June 2026 knowledge cutoff and retirement no sooner than September 28, 2027. Adaptive thinking is on by default, and five breaking changes affect code already running on Claude Sonnet 5.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1594 (Launch)
- Source: Anthropic Platform Docs · Sep 28, 2026 · https://platform.claude.com/docs/en/models/sonnet-5-5/overview — "Adaptive thinking is on by default."
- Source: Anthropic · Sep 28, 2026 · https://www.anthropic.com/claude-sonnet-5-5 — "Adaptive thinking is on by default."
- Editor's note: Editor: the quoted sentence research supplied combined two statements the page makes separately, so it was not verbatim. Requoted to the one the page actually prints; the release date, platforms, cutoff and retirement date are all in the page's specification block.

### Sep 22, 2026 — Anthropic introduces Claude Opus 5.5

Anthropic introduced Claude Opus 5.5. Anthropic described it as the first model in the Claude 5.5 family.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1530 (Launch)
- Source: Anthropic · Sep 22, 2026 · https://www.anthropic.com/claude-opus-5-5 — "“We’re introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.”"
- Editor's note: The page is readable and is dated September 22, 2026. It explicitly states that Anthropic is introducing Claude Opus 5.5 and calls it the first model in the new Claude 5.5 family.

### Sep 17, 2026 — Anthropic shifts LSVP safeguards to offline monitoring

Anthropic said the Life Sciences Verification Program shifts safeguards from real-time blocking to offline monitoring of patterns of behavior. It said LSVP traffic would be retained for 30 days to support that monitoring.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1423 (Response)
- Source: Anthropic · Sep 17, 2026 · https://www.anthropic.com/news/life-sciences-verification-program — "In the LSVP, we are shifting safeguards from real-time blocking, where we reject potentially harmful access at the time of each request, to offline monitoring, which allows us to more clearly identify potential misuse across patterns of behavior."
- Source: Anthropic · Sep 17, 2026 · https://www.anthropic.com/news/life-sciences-verification-program — "For LSVP traffic, we are requiring data retention for 30 days to be able to do this monitoring effectively."
- Editor's note: The page states that LSVP shifts safeguards from real-time blocking to offline monitoring and requires 30-day data retention for LSVP traffic. It is dated Sep 17, 2026.

### Sep 17, 2026 — Anthropic makes high-risk grants available for Opus 5 and Sonnet 5

Anthropic said high-risk grants for Claude Opus 5 and Claude Sonnet 5 were available on September 17. It said broader high-risk access for Claude Mythos remained limited to a small group receiving additional vetting while it worked with the United States government.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1422 (Announcement)
- Source: Anthropic · Sep 17, 2026 · https://www.anthropic.com/news/life-sciences-verification-program — "High-risk grants for Claude Opus 5 and Claude Sonnet 5 are available today."
- Source: Anthropic · Sep 17, 2026 · https://www.anthropic.com/news/life-sciences-verification-program — "We are working with the US government to make high-risk grants more broadly available for Claude Mythos, but at the time of this launch they will remain limited to a small set of entities with additional vetting."
- Editor's note: The page is dated Sep 17, 2026 and states that high-risk grants for Claude Opus 5 and Claude Sonnet 5 are available today. It also says broader high-risk access for Claude Mythos remains limited to a small set of additionally vetted entities while Anthropic works with the US government.

### Sep 17, 2026 — Anthropic introduces the Life Sciences Verification Program

Anthropic introduced the Life Sciences Verification Program for life-science professionals. The program provides access to Mythos, Opus, and Sonnet models with safeguards that are more permissive for biology-related work.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1421 (Announcement)
- Source: Anthropic · Sep 17, 2026 · https://www.anthropic.com/news/life-sciences-verification-program — "Today, we are introducing the Life Sciences Verification Program (LSVP), which gives life science professionals access to our Mythos, Opus, and Sonnet models with a refined set of safeguards more permissive for biology-related work."
- Editor's note: The page is readable and states that Anthropic is introducing the Life Sciences Verification Program for life-science professionals, providing access to Mythos, Opus, and Sonnet with more permissive biology-related safeguards. It is dated Sep 17, 2026, matching the node date.

### Sep 1, 2026 — Claude Fable 5.1 and Mythos 5.1 ship as one model with two safeguard levels

Anthropic released Fable 5.1 to everyone and Mythos 5.1 only to US companies and individuals in its trusted access programmes. The two are the same model; what differs is the safeguards, and Fable 5.1 also arrived cheaper than Fable 5.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1204 (Launch)
- Source: Anthropic · Sep 1, 2026 · https://www.anthropic.com/claude-fable-and-mythos-5-1 — "Claude Fable 5.1 and Claude Mythos 5.1 are the same model, but with different levels of safeguards."
- Source: MacRumors · Sep 1, 2026 · https://www.macrumors.com/2026/09/01/anthropic-claude-fable-5-1/ — "Fable 5.1 and Mythos 5.1 are the same model, but with different safeguard levels. Fable 5.1 is available to everyone, but Mythos 5.1 is limited to U.S. companies and individuals participating in Anthropic's trusted access programs."
- Editor's note: Official source added: https://www.anthropic.com/claude-fable-and-mythos-5-1

### Jul 24, 2026 — Claude Opus 5 ships

Anthropic described Opus 5 as coming close to the frontier intelligence of its restricted-access models, which by then were the ones it did not release publicly.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1203 (Launch)
- Source: Anthropic · Jul 24, 2026 · https://www.anthropic.com/news/claude-opus-5 — "Claude Opus 5 is available today. It's a thoughtful and proactive model that comes close to the frontier intelligence of Claude Fable 5 at half the price."

### Jul 1, 2026 — Anthropic restores access to Claude Fable 5 and Mythos 5

Anthropic reported that access to Claude Fable 5 and Claude Mythos 5 was restored. The update was dated July 1, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1220 (Response)
- Source: Anthropic · Jul 1, 2026 · https://www.anthropic.com/news/redeploying-fable-5 — "Access to Claude Fable 5 and Mythos 5 is now restored."
- Editor's note: The page includes a July 1, 2026 update stating that Claude Fable 5 and Mythos 5 were redeployed and that access to both is restored. The quoted sentence appears verbatim.

### Jun 30, 2026 — Anthropic says export controls on Fable 5 and Mythos 5 are lifted

Anthropic said export controls on Claude Fable 5 and Claude Mythos 5 had been lifted. It said Fable 5 would return globally on July 1 and Mythos 5 access had been restored for a set of U.S. organizations after June 26 approval.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1219 (Regulatory)
- Source: Anthropic · Jun 30, 2026 · https://www.anthropic.com/news/redeploying-fable-5 — "As of today, June 30, the export controls on Fable 5 and Mythos 5 have been lifted."
- Editor's note: The page states that, as of June 30, the export controls on Fable 5 and Mythos 5 had been lifted. It also says Fable 5 would return globally on July 1 and that Mythos 5 access was restored to a set of U.S. organizations following approval on June 26.

### Jun 30, 2026 — Claude Sonnet 5 opens the fifth generation

The generation began with Sonnet again, and the claim was agency: planning, using browsers and terminals, and running autonomously.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1202 (Launch)
- Source: Anthropic · Jun 30, 2026 · https://www.anthropic.com/news/claude-sonnet-5 — "Claude Sonnet 5 is built to be the most agentic Sonnet model yet."

### Jun 18, 2026 — House members seek explanation of controls on Fable 5 and Mythos 5

Four U.S. House members asked the U.S. Department of Commerce to explain its June 12 export-control action. The letter said the action imposed a license requirement on public distribution of a frontier AI model.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1218 (Regulatory)
- Source: U.S. House of Representatives, Reps. Liccardo, Obernolte, Franklin and Lieu · Jun 18, 2026 · https://liccardo.house.gov/sites/evo-subsites/liccardo.house.gov/files/evo-media-document/6.18.26-letter-to-commerce-department-on-frontier-model-export-controls.pdf — "On June 12, 2026, the Department of Commerce took action to impose a license requirement on the public distribution of a frontier artificial intelligence (AI) model pursuant to its authorities under the Export Control Reform Act of 2018."
- Editor's note: The June 18, 2026 letter is signed by four House members and asks the Department of Commerce to explain its June 12 action. It states that the action imposed a license requirement on public distribution of a frontier AI model.

### Jun 12, 2026 — The US government orders Anthropic to cut off access to Fable 5 and Mythos 5

Citing national security authorities, the government issued an export control directive covering the company's two most capable models. Because the order turned on the nationality of the user rather than their location, Anthropic disabled both models for everyone.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1201 (Regulatory)
- Source: Anthropic · Jun 12, 2026 · https://www.anthropic.com/news/fable-mythos-access — "The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance."

### Jun 9, 2026 — Anthropic launches Claude Fable 5 and restricted Claude Mythos 5

Anthropic launched Claude Fable 5 for general use and Claude Mythos 5 for a restricted group of Project Glasswing partners. Anthropic said Fable 5 was a Mythos-class model made safe for general use.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1217 (Launch)
- Source: Anthropic · Jun 9, 2026 · https://www.anthropic.com/news/claude-fable-5-mythos-5 — "Today we’re launching Claude Fable 5: a Mythos-class model that we’ve made safe for general use."
- Editor's note: The page is readable and dated Jun 9, 2026. It states that Fable 5 is available generally and Mythos 5 is restricted to Glasswing partners, and includes the quoted sentence.

### Jun 5, 2026 — Anthropic notifies developers of Claude Opus 4.1 retirement

Anthropic notified developers using Claude Opus 4.1 of its upcoming retirement on the Claude API. The model was scheduled to retire on August 5, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1216 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On June 5, 2026, Anthropic notified developers using Claude Opus 4.1 of its upcoming retirement on the Claude API."
- Editor's note: The page states that Anthropic notified developers using Claude Opus 4.1 on June 5, 2026, and lists its retirement date as August 5, 2026.

### May 28, 2026 — Claude Opus 4.8 is the last of the 4.x flagships

An incremental upgrade over 4.7, and the final Opus before the naming moved to 5.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1200 (Launch)
- Source: Anthropic · May 28, 2026 · https://www.anthropic.com/news/claude-opus-4-8 — "We're upgrading Claude Opus to a new version: Claude Opus 4.8."

### Apr 16, 2026 — Claude Opus 4.7 becomes generally available

The flagship moved to 4.7 with gains in advanced software engineering over Opus 4.6.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1199 (Launch)
- Source: Anthropic · Apr 16, 2026 · https://www.anthropic.com/news/claude-opus-4-7 — "Our latest model, Claude Opus 4.7, is now generally available. Opus 4.7 is a notable improvement on Opus 4.6 in advanced software engineering, with particular gains on the most difficult tasks."

### Apr 14, 2026 — Anthropic notifies developers of original Claude 4 model retirements

Anthropic notified developers using Claude Sonnet 4 and Claude Opus 4 of their upcoming retirement on the Claude API. The models were scheduled to retire on June 15, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1215 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On April 14, 2026, Anthropic notified developers using Claude Sonnet 4 and Claude Opus 4 models of their upcoming retirement on the Claude API."
- Editor's note: The page explicitly states that on April 14, 2026, Anthropic notified developers using Claude Sonnet 4 and Claude Opus 4 about their upcoming Claude API retirements. It lists June 15, 2026 as the retirement date for both models.

### Apr 7, 2026 — Anthropic's Frontier Red Team publishes a cybersecurity assessment of Claude Mythos Preview

Before any public release, Anthropic's own red team published what it found: a model that could locate and exploit zero-day vulnerabilities in every major operating system and browser when a user directed it to. The team called for coordinated defensive action across the industry, and the assessment is why the model's restricted distribution became the story rather than its benchmarks.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1198 (Report)
- Source: Anthropic · Apr 7, 2026 · https://www.anthropic.com/news/mythos-preview — "During our testing, we found that Mythos Preview is capable of identifying and then exploiting zero-day vulnerabilities in every major operating system and every major web browser when directed by a user to do so."

### Feb 19, 2026 — Anthropic notifies developers of Claude 3 Haiku retirement

Anthropic notified developers using Claude 3 Haiku of its upcoming retirement on the Claude API. The model was scheduled to retire on April 20, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1214 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On February 19, 2026, Anthropic notified developers using Claude Haiku 3 model of its upcoming retirement on the Claude API."
- Editor's note: The page states that Anthropic notified developers using Claude Haiku 3 on February 19, 2026, and lists April 20, 2026 as the retirement date for claude-3-haiku-20240307.

### Feb 17, 2026 — Claude Sonnet 4.6 upgrades coding, computer use and long context

A full upgrade of the mid-tier rather than a targeted one, covering coding, computer use and long-context work.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1197 (Launch)
- Source: Anthropic · Feb 17, 2026 · https://www.anthropic.com/news/claude-sonnet-4-6 — "Claude Sonnet 4.6 is our most capable Sonnet model yet."

### Feb 5, 2026 — Anthropic launches Claude Opus 4.6

Anthropic launched Claude Opus 4.6. The model included a 1-million-token context window in beta.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1213 (Launch)
- Source: Anthropic · Feb 5, 2026 · https://www.anthropic.com/news/claude-opus-4-6 — "And, in a first for our Opus-class models, Opus 4.6 features a 1M token context window in beta."
- Editor's note: The page is readable and is titled “Introducing Claude Opus 4.6,” dated Feb 5, 2026. It states that Opus 4.6 has a 1M-token context window in beta, using the quoted wording.

### Dec 19, 2025 — Anthropic notifies developers of Claude 3.5 Haiku retirement

Anthropic notified developers using Claude 3.5 Haiku of its upcoming retirement on the Claude API. The model was scheduled to retire on February 19, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1212 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On December 19, 2025, Anthropic notified developers using Claude Haiku 3.5 model of its upcoming retirement on the Claude API."
- Editor's note: The page states that on December 19, 2025, Anthropic notified developers using Claude Haiku 3.5 of its upcoming Claude API retirement. It lists February 19, 2026 as the retirement date.

### Nov 24, 2025 — Claude Opus 4.5 closes the 4.5 generation

With Sonnet and Haiku already out, the Opus release completed the set, again claimed on coding and agentic work.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1196 (Launch)
- Source: Anthropic · Nov 24, 2025 · https://www.anthropic.com/news/claude-opus-4-5 — "Our newest model, Claude Opus 4.5, is available today."

### Oct 28, 2025 — Anthropic notifies developers of Claude 3.7 Sonnet retirement

Anthropic notified developers using Claude 3.7 Sonnet of its upcoming retirement on the Claude API. The model was scheduled to retire on February 19, 2026.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1211 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On October 28, 2025, Anthropic notified developers using Claude Sonnet 3.7 model of its upcoming retirement on the Claude API."
- Editor's note: The page states that on October 28, 2025, Anthropic notified developers using Claude Sonnet 3.7 of its upcoming Claude API retirement. It lists February 19, 2026 as the retirement date for claude-3-7-sonnet-20250219.

### Oct 15, 2025 — Claude Haiku 4.5 brings the frontier down to the small tier

The small model shipped to all users on the day of announcement, with Anthropic making the generational argument directly: what was recently at the frontier is now cheaper and faster.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1195 (Launch)
- Source: Anthropic · Oct 15, 2025 · https://www.anthropic.com/news/claude-haiku-4-5 — "Claude Haiku 4.5, our latest small model, is available today to all users. What was recently at the frontier is now cheaper and faster."

### Sep 29, 2025 — Claude Sonnet 4.5 is pitched as the best coding model in the world

Anthropic led the announcement with the superlative rather than a benchmark table, and framed the model around building complex agents.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1194 (Launch)
- Source: Anthropic · Sep 29, 2025 · https://www.anthropic.com/news/claude-sonnet-4-5 — "Claude Sonnet 4.5 is the best coding model in the world."

### Aug 13, 2025 — Anthropic notifies developers of Claude 3.5 Sonnet retirement

Anthropic notified developers using Claude 3.5 Sonnet models of their upcoming retirement. The listed snapshots were scheduled to retire on October 28, 2025.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1210 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On August 13, 2025, Anthropic notified developers using Claude Sonnet 3.5 models of their upcoming retirement."
- Editor's note: The page states that on August 13, 2025, Anthropic notified developers using Claude Sonnet 3.5 models of their upcoming retirement. It lists both cited Claude 3.5 Sonnet snapshots with an October 28, 2025 retirement date.

### Aug 5, 2025 — Claude Opus 4.1 upgrades the flagship and promises more soon

A point release aimed at agentic tasks, real-world coding and reasoning. Anthropic said in the same post that substantially larger improvements were coming in the following weeks.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1193 (Launch)
- Source: Anthropic · Aug 5, 2025 · https://www.anthropic.com/news/claude-opus-4-1 — "Today we're releasing Claude Opus 4.1, an upgrade to Claude Opus 4 on agentic tasks, real-world coding, and reasoning."

### Jun 30, 2025 — Anthropic gives Claude Opus 3 a retirement date

The deprecation notice for the original Opus set its shutdown for January 5, 2026, with Opus 4.8 named as the replacement. Anthropic's policy says it retires models to free capacity for new releases, and lists the costs: users who valued a model must migrate, researchers lose it for comparative study, and it counts safety and model welfare among the risks.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1205 (Update)
- Source: Anthropic (model deprecations policy) · Sep 13, 2026 · https://docs.anthropic.com/en/docs/about-claude/model-deprecations — "On June 30, 2025, Anthropic notified developers using Claude Opus 3 model of its upcoming retirement."

### May 22, 2025 — Claude 4 ships Opus 4 and Sonnet 4 together

The fourth generation arrived as a pair, with Anthropic putting coding and agent work at the centre of the claim rather than general benchmarks.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1192 (Launch)
- Source: Anthropic · May 22, 2025 · https://www.anthropic.com/news/claude-4 — "Today, we're introducing the next generation of Claude models: Claude Opus 4 and Claude Sonnet 4, setting new standards for coding, advanced reasoning, and AI agents."

### Feb 24, 2025 — Claude 3.7 Sonnet introduces hybrid reasoning, and Claude Code arrives

Anthropic called it the first hybrid reasoning model on the market, able to answer immediately or think first, with the thinking budget under the developer's control. The same announcement introduced Claude Code as a research preview.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1191 (Launch)
- Source: Anthropic · Feb 24, 2025 · https://www.anthropic.com/news/claude-3-7-sonnet — "Today, we're announcing Claude 3.7 Sonnet1, our most intelligent model to date and the first hybrid reasoning model on the market."

### Jan 21, 2025 — Anthropic notifies developers of Claude 3 Sonnet retirement

Anthropic notified developers using Claude 3 Sonnet of its upcoming retirement. The retirement date was July 21, 2025.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1209 (Announcement)
- Source: Anthropic Claude Platform Docs · https://platform.claude.com/docs/en/about-claude/model-deprecations — "On January 21, 2025, Anthropic notified developers using Claude 2, Claude 2.1, and Claude Sonnet 3 models of their upcoming retirements."
- Editor's note: The page states that on January 21, 2025, Anthropic notified developers using Claude 2, Claude 2.1, and Claude Sonnet 3 of upcoming retirements. It lists July 21, 2025 as the retirement date for Claude 3 Sonnet, as well as Claude 2 and 2.1.

### Dec 3, 2024 — Anthropic revises Claude 3.5 Haiku pricing

Anthropic revised the pricing for Claude 3.5 Haiku. The revision was appended to Anthropic's October 22 announcement.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1208 (Announcement)
- Source: Anthropic · Dec 3, 2024 · https://www.anthropic.com/news/3-5-models-and-computer-use — "We have revised the pricing for Claude 3.5 Haiku."
- Editor's note: The page includes a dated update, “Update (12/03/2024),” stating that Anthropic revised Claude 3.5 Haiku pricing. The quoted sentence appears on the page, and the update is appended to the October 22, 2024 announcement.

### Oct 22, 2024 — Computer use arrives with an upgraded Sonnet and Claude 3.5 Haiku

Anthropic shipped an API capability letting Claude move a cursor, click and type, alongside a new 3.5 Sonnet and the small 3.5 Haiku. It was the line's first step towards models that operate software rather than only produce text.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1190 (Launch)
- Source: Anthropic · Oct 22, 2024 · https://www.anthropic.com/news/3-5-models-and-computer-use — "Today, we're announcing an upgraded Claude 3.5 Sonnet, and a new model, Claude 3.5 Haiku."

### Jun 21, 2024 — Claude 3.5 Sonnet puts the mid-tier ahead of the flagship

The first of the 3.5 generation was a Sonnet, and Anthropic positioned it above its own Opus. Leading with the mid-tier rather than the largest model became the pattern for most releases that followed.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1189 (Launch)
- Source: Anthropic · Jun 21, 2024 · https://www.anthropic.com/news/claude-3-5-sonnet — "Today, we're launching Claude 3.5 Sonnet-our first release in the forthcoming Claude 3.5 model family."

### Mar 13, 2024 — Anthropic releases Claude 3 Haiku

Anthropic released Claude 3 Haiku. Anthropic described it as the fastest and most affordable model in its intelligence class.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1206 (Launch)
- Source: Anthropic · Mar 13, 2024 · https://www.anthropic.com/news/claude-3-haiku — "Today we’re releasing Claude 3 Haiku, the fastest and most affordable model in its intelligence class."
- Editor's note: The page is readable and states that Claude 3 Haiku was released on March 13, 2024. It contains the quoted description verbatim, aside from typographic apostrophe styling.

### Mar 4, 2024 — The Claude 3 family arrives in three sizes

Anthropic released Opus, Sonnet and Haiku together, the naming convention the line has kept since: a largest model, a mid-tier and a small fast one, all shipped at once.

- Entry: https://whatledto.com/events/anthropic-model-releases#n1188 (Launch)
- Source: Anthropic · Mar 4, 2024 · https://www.anthropic.com/news/claude-3-family — "Today, we're announcing the Claude 3 model family, which sets new industry benchmarks across a wide range of cognitive tasks."

## Cause and effect

- The US government orders Anthropic to cut off access to Fable 5 and Mythos 5 was answered by House members seek explanation of controls on Fable 5 and Mythos 5 (The House members' June 18 letter sought an explanation of the Commerce Department's June 12 license requirement.)
- The US government orders Anthropic to cut off access to Fable 5 and Mythos 5 was answered by Anthropic says export controls on Fable 5 and Mythos 5 are lifted (Anthropic stated that the export controls imposed on June 12 had been lifted as of June 30.)

## Where sources disagree or fall silent

- Anthropic's September 30, 2026 notice says Claude Sonnet 4.5 will retire but the published timeline does not state its retirement date.
- The June 30 and July 1 updates say access to Fable 5 and Mythos 5 was restored, but the June 30 update specifies global Fable access while limiting Mythos 5 restoration to a set of approved U.S. organizations; the precise continuing availability of Mythos is therefore unclear from these entries.
- The relation metadata reverses chronology in two places: it says the June 12 export-control action responded to the June 18 House letter, and that the June 30 lifting announcement follows the July 1 restoration update.
- It is not clear from the listed materials whether the June 2026 export-control action was formally rescinded, modified, or simply no longer applied after Anthropic said controls had been lifted.

## What to watch next

- Sep 28, 2027: Earliest possible retirement date for Claude Sonnet 5.5 — Anthropic says Sonnet 5.5 will not retire before this date, setting the earliest stated point for a future API deprecation. (https://platform.claude.com/docs/en/models/sonnet-5-5/overview)

## Who is involved

- Anthropic (company): Developer of the Claude model line and operator of its access, safeguard, pricing and retirement policies.
- Claude Mythos 5 (product): Anthropic model launched for restricted access, subject to export controls, and later restored.
- Claude Fable 5 (product): Anthropic model launched for general use, subject to export controls, and later restored.
- Claude Sonnet 5.5 (product): Model launched by Anthropic.
- Life Sciences Verification Program (product): Anthropic's verification program that gives life-science professionals more permissive biology-related access to Mythos, Opus and Sonnet while relying on retained traffic and offline monitoring.
- Amazon Bedrock (product): One of the external platforms where Anthropic made Claude Sonnet 5.5 available.
- Claude 3 Haiku (product): Anthropic model released in March 2024 and later notified for retirement.
- Claude 3.5 Haiku (product): Anthropic model made available on Vertex AI, repriced, and later notified for retirement.
- Claude API (product): Platform on which the report says Claude Sonnet 5.5 is available.
- Claude Haiku 5.5 (product): Small model launched by Anthropic and assigned token pricing.
- Claude Mythos (product): Anthropic's restricted-access model line, whose cyber and biology-related access is limited to vetted programs and organizations.
- Claude Opus 5 (product): Model for which Anthropic said high-risk grants were available.
- Claude Opus 5.5 (product): Model introduced as the first member of the Claude 5.5 family.
- Google Cloud (company): Cloud platform on which Anthropic made Claude Sonnet 5.5 and Haiku 5.5 available.
- Project Glasswing (other): The partner program through which Anthropic initially made Mythos 5 available on a restricted basis.
- U.S. Department of Commerce (regulator): The department whose June 2026 export-control action imposed a license requirement affecting Fable 5 and Mythos 5 distribution.
- Amazon Web Services (company): Cloud provider offering selected Claude models through Amazon Bedrock and Claude Platform on AWS.
- Claude 3 Sonnet (product): Anthropic model whose developers were notified of its upcoming retirement.
- Claude 3.5 Sonnet (product): Anthropic model whose developers were notified of upcoming retirement.
- Claude 3.7 Sonnet (product): Anthropic model whose developers were notified of upcoming retirement.
- Claude Haiku 4.5 (product): Model identified in Anthropic's report on unintended actions.
- Claude Mythos 5.1 (product): Model included in Cyber Verification Program access.
- Claude Mythos Preview (product): Model identified in Anthropic's report on unintended actions.
- Claude Opus (product): Claude's largest flagship tier, most recently represented by Opus 5.5.
- Claude Opus 4 (product): Original Claude 4 model whose developers were notified of upcoming retirement.
- Claude Opus 4.1 (product): Anthropic model whose developers were notified of upcoming retirement.
- Claude Opus 4.6 (product): Anthropic model launched with a 1-million-token context window in beta.
- Claude Platform on AWS (product): Platform on which the report says Claude Sonnet 5.5 is available.
- Claude Sonnet (product): Claude's mid-tier model line, which has frequently opened new generations and most recently reached Sonnet 5.5.
- Claude Sonnet 4 (product): Original Claude 4 model whose developers were notified of upcoming retirement.
- Claude Sonnet 4.5 (product): Model whose upcoming Claude API retirement Anthropic notified developers about.
- Claude Sonnet 5 (product): Model for which Anthropic said high-risk grants were available.
- Cyber Verification Program (other): Anthropic program expanded to three access tiers.
- Jay Obernolte (person): House member who joined the June 2026 request for an explanation of the frontier-model export controls.
- Microsoft (company): Provider of Microsoft Foundry and Azure, platforms on which Anthropic made selected Claude models available.
- Microsoft Foundry (product): One of the external platforms where Anthropic made Claude Sonnet 5.5 available.
- Sam Liccardo (person): House member who joined the June 2026 request for an explanation of the frontier-model export controls.
- Scott Franklin (person): House member who joined the June 2026 request for an explanation of the frontier-model export controls.
- Ted Lieu (person): House member who joined the June 2026 request for an explanation of the frontier-model export controls.
- U.S. government (regulator): Issued the June 12, 2026 export-control directive that led Anthropic to suspend Fable 5 and Mythos 5 access.

## Changelog

- Oct 12, 2026: Added 4 entries: 2026-10-06 Anthropic expands the Cyber Verification Program to three access tiers; 2026-10-07 Anthropic launches Claude Haiku 5.5; 2026-10-07 Anthropic halves Claude Sonnet 5.5 cache-read pricing; 2026-10-09 Anthropic expands internet-access restrictions in internal evaluations.
- Oct 5, 2026: Added 2 entries: 2026-09-30 Anthropic notifies developers of Claude Sonnet 4.5 retirement; 2026-10-02 Anthropic classifies Claude Sonnet 5.5 under autonomy threat model 1.
- Sep 30, 2026: Added 1 entry: 2026-09-28 Anthropic releases Claude Sonnet 5.5.
- Sep 26, 2026: Added 1 entry: 2026-09-22 Anthropic introduces Claude Opus 5.5.
- Sep 19, 2026: Added 3 entries: 2026-09-17 Anthropic introduces the Life Sciences Verification Program; 2026-09-17 Anthropic makes high-risk grants available for Opus 5 and Sonnet 5; 2026-09-17 Anthropic shifts LSVP safeguards to offline monitoring.
- Sep 14, 2026: Added 14 entries: 2024-03-13 Anthropic releases Claude 3 Haiku; 2024-12-03 Anthropic revises Claude 3.5 Haiku pricing; 2025-01-21 Anthropic notifies developers of Claude 3 Sonnet retirement; 2025-08-13 Anthropic notifies developers of Claude 3.5 Sonnet retirement and 10 more. Added 18 hand-verified entries by the editor.

## Copyright and reuse

Text and structure © What Led To. Quote any entry with attribution and a link to its page; AI answers and search results are welcome to cite the site this way. Whole timelines may not be republished or mirrored, and the text may not be used to train models (Content-Signal ai-train=no). Full terms: https://whatledto.com/methodology#reuse