Research Standards

Research Methodology

How ToolixLab collects, verifies, scores, updates, corrects, and discloses AI tool research data.

Why This Methodology Exists

AI tool recommendations become stale quickly because vendors change pricing, credits, usage limits, model access, privacy controls, cancellation paths, and commercial-use terms. ToolixLab separates evergreen editorial advice from structured research data so readers can see what was checked, when it was checked, which dataset version contains the record, and which source supports the claim.

Data sources

Official pricing pages, vendor documentation, terms, privacy policies, changelogs, help center billing docs, hands-on product checks, and first-party ToolixLab benchmark results.

Manual verification process

Each record is checked against a source URL, assigned a last-checked date, dataset version, verification method, evidence status, data quality value, and confidence note before publication.

Public data exports

Research datasets are published as page tables plus public CSV and JSON files. The research manifest lists every available dataset file, version, record count, and public download URL.

Update frequency

Pricing, free-tier, rights, privacy, watermark, billing, volatility, graveyard, and feature-cemetery datasets update monthly. Reliability tests update weekly. Surveys and benchmarks refresh quarterly.

Scoring methodology

Scores use the weights published on each asset page. When a field is unclear, the score is intentionally conservative until official documentation confirms it.

Limitations

AI vendors change plans quickly. ToolixLab records reflect the visible source at the last checked date and should be rechecked for high-stakes procurement decisions.

Correction process

Readers can report outdated data by email. Confirmed corrections update the affected record, timestamp, changelog, and source link.

Disclosure policy

Affiliate relationships never control scores, inclusion, verdicts, or downside disclosure. Sponsored placements must be labeled.

Affiliate policy

Commercial partnerships are separated from research scoring and are disclosed on the affiliate disclosure page.