Research Data Dictionary

The product or AI tool being tracked.

The primary software category used for filtering and comparison.

The vendor homepage or official product URL.

The source used to verify the record. Prefer official vendor URLs.

The date ToolixLab last checked the source.

The date ToolixLab last changed the public record.

Current record status such as active, inactive, shut down, rebranded, or acquired.

Stable internal record identifier.

Public citation identifier for referencing a specific dataset row.

Dataset release identifier, usually the monthly, weekly, quarterly, or benchmark refresh period.

Date the source URL was accessed for verification.

Optional archived copy used for historical comparison.

How the claim was checked: official source, hands-on test, manual review, survey, benchmark, or pending.

Whether the record is verified, needs refresh, pending, or disputed.

High, medium, low, or pending confidence based on source clarity and verification depth.

Plain-language caveat that explains how strongly the record should be trusted.

Composite trust score across pricing, free-plan usefulness, privacy, commercial rights, lock-in, billing friction, reliability, and volatility.

How useful the free plan is after usage limits, credit cards, watermarks, exports, commercial use, and core features.

Risk score for exports, data ownership, API availability, cancellation clarity, portability, free-plan dependency, and shutdown risk.

Friction score for trials, credit cards, cancellation path, refunds, hidden fees, and auto-renewal clarity.

How often pricing, limits, terms, access, branding, features, or availability change.

Observed generation success, speed, export success, credits consumed, error rate, and quality pass/fail.