Dedicated deployments
These rules apply to dedicated model deployments, whether the workers are always on (warm) or scale to zero and burst. A deployment may use either mode, so both modes follow this section. The Client must be at least 18 years old. The Client must not submit, request, generate, store, or distribute:- Sexual or pornographic content, graphic adult content, or other adult products, including material that promotes those activities.
- Content that exploits or harms a minor, exposes a minor to inappropriate content, or solicits a minor into sexual activity. This includes child sexual abuse material. SynapsAI will suspend the account and report apparent violations as the law requires.
- Content that is obscene, lewd, violent, harassing, threatening, or otherwise objectionable, including content that abuses a person or promotes violence against a person or a class of people.
- Intoxicants, illegal drugs or other illegal products, alcoholic beverages, or games of chance.
- Content that infringes intellectual property, violates privacy or publicity rights, or is fraudulent or misleading.
- Spam or other unsolicited promotional material.
- Material in which the Client impersonates SynapsAI, its staff, or another person.
- Malware, or any attempt to gain unauthorized access, probe or scrape the service, disrupt other tenants, or run a denial-of-service attack.
- A workload located in, or exported or re-exported to, a country or region subject to comprehensive sanctions, a prohibited party, or the Government of Venezuela.
Serverless catalog
These rules apply to every call to a model in the shared serverless catalog (serverless-catalog:).
The Client warrants that each input, and any content the Client asks the model to produce, excludes:
- Nudity or other sexually suggestive content.
- Hate speech, threats, or direct attacks on an individual or group.
- Content that is abusive, harassing, tortious, defamatory, vulgar, obscene, libelous, or invasive of privacy.
- Sexist content, or content that discriminates on the basis of race, ethnicity, or another protected ground.
- Self-harm content or excessive violence.
- Fake or impostor profiles, illegal content, or content in furtherance of harmful or illegal activities.
- Malicious programs or code.
- Any person’s personal information without that person’s consent.
- Personal information of children under 13, or under the applicable age of digital consent.
- Spam, or machine-generated content sent as an unsolicited message.
- Information a government body deems sensitive or classified.
- Payment card numbers, CVVs, or other cardholder data.
- Content that infringes a third party’s intellectual property, or that the Client lacks the rights to submit.