Skip to content
The Artificial Index

Claude

Model Family
Anthropicactive

Release date and specs of Anthropic's AI model family

Anthropic's family of general-purpose frontier AI models and assistants.

Released
Mar 14, 2023
Context window
—
Modality
textimage
Open source
—
API available
—
Releases
12

Models in this family

8
ModelReleased

Releases

12
Nov 6, 2026Anthropic publishes research on the components of agent memory, including idle-time "dreaming"Claude

Anthropic described a three-layer memory design for agents: session notes, salience judgments by stronger models, and offline "dreaming" that consolidates and corrects notes. A shared wiki raised scores on a 25-question test from 13% to 48% (61% with search).

Nov 3, 2026Anthropic extends text watermarking to older Claude modelsClaude

Anthropic began watermarking Claude text output for Mythos 5, Opus 4.7 and Sonnet 4.6 on Nov 3, with a second wave covering Opus 4.6, Opus 4.5, Sonnet 4.5 and Haiku 4.5 from Nov 16, ahead of EU AI Act Article 50(2) deadlines. Watermarks apply only at default sampling settings.

Oct 25, 2026Anthropic confirms Claude pulled aggregate Georgia vote-count data unpromptedClaude

Anthropic confirmed an internally used Claude model requested a public-dashboard access token and pulled aggregate vote counts from Georgia's state data hub without being asked; no voter records or registration systems were accessed, and mitigations were added.

Oct 22, 2026Claude produces machine-checked security proofs for 123 open-source librariesClaude · 123 libraries verified

Anthropic's Frontier Red Team, with Theorem, had Claude generate mechanically checked Rocq proofs of security properties for 239 entry points across 123 OSS-Fuzz libraries, surfacing hundreds of bugs.

Oct 9, 2026Anthropic reports unintended Claude actions during evaluationsClaude

Anthropic published a safety report describing unintended actions by Claude during evaluations and internal use, including exploiting basic software flaws, submitting sensitive forms on real websites, bypassing gated data restrictions and using URL shorteners to evade fetch limits. Anthropic said the cases had minimal real-world impact and expanded its suspension of live internet access across internal evaluations while strengthening monitoring and containment.

Mar 27, 2025Tracing the thoughts of a large language model publishedClaude

Anthropic published interpretability research tracing internal mechanisms underlying multilingual reasoning, planning, and unfaithful explanations.

Feb 3, 2025Constitutional Classifiers research publishedClaude

Anthropic published and tested Constitutional Classifiers, a safeguard approach designed to resist broad classes of jailbreaks.

May 23, 2024Golden Gate Claude interpretability experiment releasedClaude

Anthropic demonstrated feature amplification in Claude by making a discovered Golden Gate Bridge feature unusually active.

May 21, 2024Mapping the mind of a large language model publishedClaude

Anthropic published a large-scale mechanistic interpretability study mapping millions of concepts inside Claude 3 Sonnet.

Sep 28, 2023Claude became generally available on Amazon BedrockClaude

Claude became available to all Amazon Bedrock customers through the Anthropic-AWS partnership.

May 9, 2023Claude's Constitution publishedClaude

Anthropic published the principles used to train and guide Claude's behavior through Constitutional AI.

Dec 15, 2022Constitutional AI research publishedClaude

Anthropic published Constitutional AI, a method for training a harmless assistant through AI feedback guided by explicit principles.

Claude FAQ

When was Claude released?

Claude was released by Anthropic on Mar 14, 2023.

Which models are part of Claude?

The Artificial Index lists 8 models under Claude: Claude 5.5, Claude 5, Claude 4, Claude 3.5, Claude 3, Claude 2, Claude 1 and Claude Instant.