service anthropic-apivendor Anthropiccategory Model APIs
No (Favourable)Commercial products (API, Claude for Work) are not trained on unless you send feedback or opt in.
Commercial products (API, Claude for Work) are not trained on unless you send feedback or opt in.
By default, we will not use your inputs or outputs from our commercial products (e.g. Claude for Work, Anthropic API, Claude Gov, etc.) to train our models.
Deleted within 30 days, except Files API and similar features you control, ZDR agreements, Usage Policy enforcement, or legal requirements.
For Anthropic API users, we automatically delete inputs and outputs on our backend within 30 days of receipt or generation
Against the 21 tools that state a period (grey dots):
Enabled per organization by the account team. 'Covered Models' (Claude Fable 5/5.1, Mythos 5/5.1) require 30-day retention and are not available under ZDR unless Anthropic authorizes it.
Under a ZDR arrangement, Anthropic does not store customer prompts or responses at rest after the API response is returned.
Flagged content is kept far longer than the 30-day default.
We retain inputs and outputs for up to 2 years and trust and safety classification scores for up to 7 years if your chat is flagged by our automated trust and safety systems as violating our
inference_geo per request is 'global' (default) or 'us'. Workspace (storage) geo: 'us' is the only option. No APAC option documented.
Default. Inference may run in any available geography for optimal performance and availability.
ISO 27001:2022 (Information Security Management) ISO/IEC 42001:2023 (AI Management Systems) SOC 2 Type I & Type II
Self-serve: eligible orgs sign the BAA and enable HIPAA readiness in the Claude Console. HIPAA readiness is permanent once enabled.
Eligible organizations can review and execute the BAA and enable HIPAA readiness directly from the Claude Console.
Claude models are listed only with IDs for the Claude API, Amazon Bedrock, Google Cloud and Microsoft Foundry. No downloadable weights or on-prem option appears in the docs.
Each model's page lists the platforms it's available on.
Spotted something wrong or out of date? Corrections are welcome; see how to send one. How statuses are assigned.