Compute Terms aria-hidden="true"bservatory

The published contract terms of cloud infrastructure and AI model providers, archived twice daily and laid out side by side. Every value links to the source document.

ai-generated summaries · documented methodology · not legal advice

Claude

Published terms, citations, and change history.

← Back to matrix

Documents archived

Extracted terms

General contract terms

Capacity & reservation terms No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

Data use & AI training ✓ Quote verified against source

Never trains on commercial Customer Content by default; consumer tier trains unless opted out

“Anthropic may not train models on Customer Content from Services.”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Suspension rights ✓ Quote verified against source

May suspend for security risk, AUP/use-restriction violation, legal prohibition, or vendor loss; reasonable-efforts notice

“Anthropic may suspend Customer's access to any portion or all of the Services if”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Termination ✓ Quote verified against source

Either party may terminate for convenience with Notice (Anthropic gives 30 days); on termination Customer loses access

“Each party may terminate these Terms at any time for convenience with Notice”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Unilateral modification ✓ Quote verified against source

Anthropic may update terms, effective 30 days after posting/Notice; law-driven changes immediate; not retroactive

“Anthropic may update these Terms at any time, to be effective 30 days after”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Governing law & disputes ✓ Quote verified against source

California (or Ireland for EEA/UK/CH); binding arbitration yes; class-action waiver yes

“THE RIGHT TO JOIN AND PARTICIPATE IN A CLASS ACTION”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Liability caps & carveouts ✓ Quote verified against source

Cap = fees paid in prior 12 months; indemnity carved out; consequential/indirect damages waived

“limited to Fees paid by Customer for the Services in the previous 12 months”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Output IP indemnification ✓ Quote verified against source

Yes for paid use; uncapped (excluded from liability cap); conditions incl. no modification, no known infringement

“Outputs generated through such authorized use violates any third-party intellectual property right”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Model deprecation & version pinning ✓ Quote verified against source

Documented policy; at least 60 days' notice before retirement; version pinning via dated model IDs

“at least 60 days' notice before model retirement for publicly released models”
Source: Anthropic model deprecations · fetched 2026-08-13

Benchmarking & eval restrictions No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

No-compete-training clauses ✓ Quote verified against source

competing-models-only (restricts building competing product / training competing AI models)

“build a competing product or service, including to train competing AI models”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Capacity delivery remedies No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

Assignment & financing clauses ✓ Quote verified against source

Consent required, except Anthropic may assign to affiliate or in sale of business; no financing/step-in language

“Anthropic may assign its rights and delegate its obligations to an affiliate or as…”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Service level (SLA) terms

Availability definition ✓ Quote verified against source

no SLA / uptime not committed

“THEIR USE WILL BE UNINTERRUPTED”
Source: Anthropic Commercial Terms of Service · fetched 2026-07-18

Credit regime No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

Claim mechanics No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

SLA exclusions No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

Data protection & privacy

International data transfer mechanism ✓ Quote verified against source

Relies on adequacy decisions and Standard Contractual Clauses; ANPD-approved SCCs for Brazil

“We rely on standard contractual clauses to transfer information”
Source: Anthropic Privacy Policy · fetched 2026-07-18

Data residency commitments ✓ Quote verified against source

No regional-processing commitment; reserves global processing/transfer to US and other countries

“your personal data is transferred to our servers in the US, or to other…”
Source: Anthropic Privacy Policy · fetched 2026-07-18

Sub-processor transparency ✓ Quote verified against source

Sub-processor list published in Trust Center; no stated advance-notice window or objection right

“All subprocessors”
Source: Anthropic Sub-processors · fetched 2026-07-21

Government access commitments ✓ Quote verified against source

Discloses to government on good-faith belief; transparency hub exists; no stated challenge/notice commitment in privacy policy

“we have a good-faith belief that disclosure is reasonably necessary”
Source: Anthropic Privacy Policy · fetched 2026-07-18

Retention & human review of content ✓ Quote verified against source

Humans may review content flagged for safety; deleted conversations removed from back-end within 30 days

“your conversations are flagged for safety review”
Source: Anthropic Privacy Policy · fetched 2026-07-18

Accountability & transparency

Prohibited high-risk uses ✓ Quote verified against source

Prohibits facial recognition surveillance, predictive policing, biometric categorization, weapons development

“facial recognition, battlefield management applications or predictive policing”
Source: Anthropic Usage Policy · fetched 2026-07-18

Model documentation commitments ✓ Quote verified against source

System cards and model reports published in Transparency Hub; not referenced contractually

“condensing key details about the models' capabilities, safety evaluations”
Source: Anthropic Transparency Hub · fetched 2026-07-25

Training data provenance statements ✓ Quote verified against source

States training on publicly available Internet info, licensed third-party datasets, opted-in user data, synthetic data

“proprietary mix of publicly available information from the Internet, public and private datasets”
Source: Anthropic model cards / Transparency Hub · fetched 2026-07-25

Appeal & redress mechanisms ✓ Quote verified against source

Privacy denials may be appealed by email; report bad outputs to usersafety; no detailed suspension-appeal timeline

“You may also have the right to appeal requests that we deny by emailing”
Source: Anthropic Privacy Policy · fetched 2026-07-18

EU AI Act role allocation No provision found

Reviewed Anthropic Commercial Terms of Service and 6 other governing documents: no provision addresses this term.

Change history

Claude: Anthropic model deprecations deprecation 2026-08-06

The status of claude-opus-4-1-20250805 was updated from "Deprecated" to "Retired" and a note was added confirming the model was retired on August 5, 2026.

Model deprecation & version pinning

AI-generated description; verify against the source before relying on it.

sourceCompare full documents side by side
Preview of changes
| claude-opus-4-1-20250805 | Deprecated | June 5, 2026 | August 5, 2026 | | claude-opus-4-1-20250805 | Retired | June 5, 2026 | August 5, 2026 | This model was retired August 5, 2026.

View prior version · View current version

Claude: Anthropic model cards / Transparency Hub ai_documentation 2026-07-25

The document's date stamp was updated and a new entry for Claude Sonnet 5 was added to the Transparency Hub, including a summary table describing its capabilities, benchmarked performance (referencing the system card's Section 8), and acceptable use policy references. A prior reference to the Claude Fable 5 & Claude Mythos 5 system card was removed.

Model documentation commitmentsProhibited high-risk uses

AI-generated description; verify against the source before relying on it.

sourceCompare full documents side by side
Preview of changes
Last updated February 20, 2026 Last updated July 23, 2026 Claude Sonnet 5 Claude Sonnet 5 Summary Table Model description Claude Sonnet 5 can make plans, use tools like browsers and terminals, and run autonomously at a level that, just a few months… See Claude Fable 5 & Claude Mythos 5 system card

View prior version · View current version

Claude: Anthropic Transparency Hub transparency_report 2026-07-25

The document was updated to add a new model entry for Claude Sonnet 5, including its model description, benchmarked capabilities reference, and acceptable use policy summary, alongside a date bump. A prior reference to a 'Claude Fable 5 & Claude Mythos 5 system card' link was removed.

Model documentation commitmentsProhibited high-risk uses

AI-generated description; verify against the source before relying on it.

sourceCompare full documents side by side
Preview of changes
Last updated February 20, 2026 Last updated July 23, 2026 Claude Sonnet 5 Claude Sonnet 5 Summary Table Model description Claude Sonnet 5 can make plans, use tools like browsers and terminals, and run autonomously at a level that, just a few months… See Claude Fable 5 & Claude Mythos 5 system card

View prior version · View current version

Claude: Anthropic model deprecations deprecation 2026-07-25

The deprecation status of Claude Mythos Preview was updated from a scheduled future retirement date (July 21, 2026) to now being marked as already deprecated. A new model entry, claude-opus-5, was added to the deprecation table listing it as Active with no scheduled deprecation date, and not sooner than July 24, 2027 for retirement. Minor formatting changes were also made to table column widths.

Model deprecation & version pinning

AI-generated description; verify against the source before relying on it.

sourceCompare full documents side by side
Preview of changes
[Claude Mythos Preview](https://anthropic.com/glasswing) (`claude-mythos-preview`) will be retired on July 21, 2026.… [Claude Mythos Preview](https://anthropic.com/glasswing) (`claude-mythos-preview`) is deprecated. To migrate to [Claude Mythos 5](https://anthropic.com/glasswing) (`claude-mythos-5`), see the [migration guide](/docs/en/about-claude/models/migration-guide#migrating-from-claude-mythos-preview). | claude-opus-5 | Active | N/A | Not sooner than July 24, 2027 | | ------------------------------- | -------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------ | -------------------------------------------------------------------------------------------------------------------------------- | | `temperature`, `top_p`, `top_k` | Deprecated (Claude Opus 4.7 and later) |… | ------------------------------- | -------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ | -------------------------------------------------------------------------------------------------------------------------------- | | `temperature`, `top_p`, `top_k` | Deprecated (Claude Opus 4.7 and later) |…

View prior version · View current version

Extracted 2026-08-13 03:34 UTC with claude-opus-4-8.