Acceptable Use Policy

This policy is part of our Terms of Service and applies to everything we operate: the diagrams.so website and app, the public API, and the official MCP server and SDKs. We keep it short so it is actually read. Violating it can lead to key revocation, removal of content, loss of access, or account termination, with unused Credits forfeited as the Terms describe.

1. Do not misuse the AI or its output

• Do not use the Services or any Output to build, train, or improve a competing AI model or diagramming service, and do not try to extract or reverse-engineer the underlying models, system prompts, or scoring logic.
• Do not generate, or attempt to generate, content that is illegal, infringing, deceptive, or harmful, or that violates the usage policies of the AI providers we route to (including the Anthropic Usage Policy and the OpenAI Usage Policies, each as updated). The following are strictly prohibited and may be reported to authorities: child sexual abuse material or any sexualization of minors; content promoting terrorism or violent extremism; instructions for building weapons or carrying out serious violence; malware, ransomware, or intrusion tooling; fraud, phishing, spam, or deceptive impersonation; and non-consensual intimate imagery or content that harasses, defames, or exposes another person's private information.
• Do not rely on Output as your sole basis for decisions in safety-critical, medical, legal, financial, or compliance contexts. Where you present Output to individuals in high-stakes domains (legal, health, financial, insurance, employment, and similar), a qualified person must review it and you must disclose that it is AI-generated. Do not use the Services for automated decisions with legal or similarly significant effects on individuals without qualified human oversight.
• Do not present Output's factual or security claims as verified without checking them, and do not present Output as wholly human-created.
• Do not remove, bypass, or falsify AI-content disclosures or machine-readable provenance markings that the Services or applicable law require.

2. Do not abuse the platform

• Do not circumvent, or attempt to circumvent, rate limits, quotas, Credit metering, or safety systems, including by rotating keys, IP addresses, or accounts, or by running scripted retry storms.
• Do not sell, share, sublicense, or transfer API keys or API access, and do not operate the Services as a bare pass-through or proxy to an AI provider. Building a real product on top of the Services is permitted and encouraged.
• Do not deliberately drive metered usage to burn Credits, whether yours or anyone else's.
• Do not create multiple accounts or identities to collect extra free Credits, evade enforcement, or exceed intended limits.
• Keep live keys secret and server-side; never commit a key to a public repository. Test-mode keys spend your real Credit balance and deserve the same care.

3. Do not attack the Services or other users

• No probing, scanning, or breaching of security or access controls, and no attempts to reach another user's diagrams or data.
• No malware, injection payloads, deliberately oversized files, or attempts to exhaust resources.
• No scraping of the Services except as the API expressly allows.
• Security researchers acting in good faith are welcome: see our Vulnerability Disclosure Policy. Report findings to security@diagrams.so rather than testing destructively.

4. Do not submit content you should not

• Only submit content you have the right to submit.
• Do not submit other people's sensitive personal data, or regulated data such as health records, payment card numbers, or government identifiers. The Services are not designed or offered for that data.
• Remember that Free plan diagrams are public. Do not put confidential or personal information in them.

5. Fair use

Published rate limits, size caps, and quotas are part of the Services; current values are in the API documentation. Sustained usage that materially exceeds your plan or degrades the experience for others may be throttled or may require an upgrade.

6. Enforcement and reporting

We investigate suspected violations and may cooperate with law enforcement. Enforcement is proportionate and normally escalates from a warning, to key revocation, to account termination, but we can move straight to revocation or termination for severe violations, security risk, or legal requirements.

To report copyright infringement or other unlawful or violating content on our public surfaces, use the process in our Copyright and Content Complaints Policy, or email dmca@diagrams.so for copyright matters and legal@diagrams.so for anything else. We review notices, act on valid ones, tell the affected user what we did and why, and terminate repeat infringers.

Questions about this policy: support@diagrams.so.