Scoring Methodology
Every tool earns four sub-scores. No paid placements.
The four sub-scores
- Voice quality. Fidelity, naturalness and pronunciation accuracy on listening samples.
- Value. Cost per minute / word at realistic usage volumes.
- UI. Onboarding, workflow, editor, API documentation.
- Maturity. Availability, uptime, history, funding, roadmap.
Overall score
The overall score is a weighted average of the four sub-scores (scale 0–5). Voice quality counts twice — it is the most important factor in production projects.
How often we update
Pricing and feature data is checked daily by our crawler and updated when anything changes. Voice-quality scores are re-evaluated quarterly when providers ship new models.
What we don't do
We do not accept payment for higher scores or better placement. Affiliate relationships do not affect ranking in comparison tables — see the disclosure page.