Smallest.ai models, access and benchmark guide
Smallest.ai model and access guide with a practical US-English evaluation brief and links to publisher materials.

👁 111 reads
A low-latency product can become expensive when the original voice reads a ZIP code incorrectly and forces the whole message to repeat. Find the Lightning Pro line in the detailed tariff and record when the first audible sample arrives.
Choosing the first job
Compare a ZIP code beginning with zero against the source script and measure startup time separately from file duration.
Models in the November 17 snapshot
| Model and track | Placement | Reported score |
|---|---|---|
| Lightning V3.1 Pro (Oct 2026) (AA provider voice) | 8 | 1271 Elo |
| Lightning V3.1 Pro (Oct 2026) (AA controlled voice) | 12 | 1101 Elo |
Release record. Publisher documentation identifies the product family. The ranked configurations include Lightning V3.1 Pro (Oct 2026).
Google · Your sourcesQuantslant higher in your Google - one clickAdd to preferred sources →Sources
- https://smallest.ai/ (accessed November 17, 2026)
- https://waves-docs.smallest.ai/ (accessed November 17, 2026)
- https://artificialanalysis.ai/text-to-speech/leaderboard/provider-voice (accessed November 17, 2026)
- https://artificialanalysis.ai/text-to-speech/leaderboard/controlled-voice (accessed November 17, 2026)
More on this from QUANTSLANT
benchmarks·Skywork AI models, access and benchmark guideSkywork AI model and access guide with a practical US-English evaluation brief and links to publisher materials.