Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
exact.works positions itself as the “contracting layer for the agentic economy.” It is not an AI tool that directly generates content; rather, it provides infrastructure for contracts, audits, escrow settlement, and dispute resolution around AI Agent service transactions. Its core components are the SAISA standard AI service agreement and Paper: these bind human-readable legal terms with machine-executable logic, creating verifiable records through a hash chain.
The platform covers the full lifecycle of AI Agent tasks. SAISA defines authorization scope, success criteria, behavioral baselines, escrow payments, and dispute pathways. Trace records every Agent action, with an emphasis on tamper resistance and auditability. APEX-BG provides real-time behavioral governance and can pause execution when it detects capabilities or behavior patterns not present in the fingerprint. Parler uses three independent evaluators—Claude, GPT-4o, and Gemini—for a three-panel-style dispute assessment, then applies confidence-weighted Bayesian aggregation to produce a probabilistic conclusion. It is worth noting that Parler’s own wording explicitly classifies this as Expert Determination, not binding arbitration.
The main materials do not disclose specific pricing, commission rates, or free trials. API examples show that the platform involves escrow amounts, service fees, and Stripe transfers. Standard limits are 100 requests/minute, Exacting 10 papers/minute, and Webhook 1000 events/hour, with higher limits available for enterprise plans. Developer support includes REST API, Webhook, Node.js/TypeScript SDK, and Python SDK. The API covers Paper creation, acceptance, disputes, verification, audit trails, and revisions.
Its strengths are clear positioning and its ability to fill common gaps in enterprise AI Agent procurement: who is responsible, how acceptance is determined, and how issues are resolved when something goes wrong. Its API and standards documentation are fairly complete, making it suitable for integration into platform-style Agent marketplaces or internal enterprise governance workflows. The limitations are also obvious: there is no public pricing, SLA, privacy compliance, data residency, or governing-law detail. Claims such as “court-admissible” and “legally enforceable” still require lawyer review and practical validation across different jurisdictions. The platform does not guarantee the quality of Agent outputs; it is responsible for recording, evaluation, and dispute interpretation.
It is best suited to buyers of enterprise-grade Agents in finance, legal, healthcare, government, and large organizations, as well as AI Providers that want to add contracting and audit capabilities to their own Agents. Ordinary individual users or those who only need chat or writing tools are not the core audience. Access from China, Chinese-language interface, RMB payments, and local compliance information have not been disclosed, so their status is currently unknown. For deployment in mainland China, local e-signature, contract management, audit log, and AI governance tools may be needed as alternatives or complements.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on exact.works official site.
exact.works is an United States AI Apps provider. TG4G tracks its product information, an overall rating of 7.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach exact.works directly.