Abacus AI
Abacus.AI · Agent platform · US · abacus.ai
services/abacus-ai.yamlVendor data-handling terms
“Such use is strictly limited to maintaining and improving the Abacus.AI platform and does not involve using personal information for developing, training, or improving any generalized or proprietary AI/ML models. / Any data exchanged with such models occurs only for real-time inference and is never stored, logged, or reused by the model provider for training or product improvement.”
“Following termination or deactivation of your account, Abacus.AI, its Clients, Affiliates, or its service providers may retain information (including your profile information) and user Content for a commercially reasonable time for backup, archival, and/or audit purposes.”
“Open-source models such as Gemma and others are used with their open-source weights and hosted by Abacus.AI entirely within secure US based servers.”
Domains and endpoints
observed means seen in Unseen deployments; vendor-documented means listed by the vendor. Vendors do not publish complete lists.
| Host | Role | Source |
|---|---|---|
| abacus.ai | app | observed |
| api.abacus.ai | api | observed |
| apps.abacus.ai | app | observed |
Assessment
Reasoning: The policy excludes training on personal information and states that third-party model providers see data only for inference, which is a stronger commitment than most aggregators make. Processing is in the United States, and content may be kept after the account closes for an unspecified "commercially reasonable time". As a multi-model front end, the service forwards prompts to several providers under its own terms rather than the user's.
Flags: GDPR Chapter V (US hosting) · Post-termination retention period unspecified · Assessed 2026-09-16. The assessment is Unseen's; the terms above are the vendor's.
Changelog
- 2026-09-16Documented. Training, retention, and hosting terms verified against the Abacus.AI privacy policy.
- 2026-09-15Listed from the Unseen catalogue with observed domains.
Corrections
Pull request on GitHub, or the form below. Changes are reviewed and recorded in the changelog.