Sign in
Pending benchmarksTracking since Nov 12, 2026

Is Tonality state of the art?Pending benchmarks.Announced Nov 12, 2026; no independent results yet.

Riverside Labs' voice model for word-level speech editing: pitch, tone, duration, pauses, energy and pronunciation, including changing, inserting or removing words in an existing recording while keeping the speaker's voice and room acoustics. It also runs on text-to-speech. Announced November 12, 2026. A free, rate-limited playground is public on the official Riverside Labs site. API keys, docs and volume pricing are offered through sales, not as a self-serve API. No public weight release is stated.

Announced facts

The details so far

Lab
Riverside Labs
Announced
Nov 12, 2026
API
Not yet
Open weights
Closed · No public weight release stated on the official product page
Type
Voice model for word-level control of pitch, tone, duration, pause, energy and pronunciation, including in-line edits that change, insert or remove words in a real recording, or generate speech from text with a reference voice (Riverside Labs product page)
Editing
Official page: edits and new words keep the speaker's voice, pacing, accent, and the acoustics of the room; the output includes the edited audio and the sample ranges that changed (Riverside Labs product page)
Hosted access
Public, rate-limited playground on the official Tonality page and at labs.riverside.com/start, described as free. API access, keys, docs and volume pricing are offered through sales, not as a public self-serve API (Riverside Labs product page)
Pricing
Official page: playground is free and rate-limited; API volume pricing is not listed and requires contacting sales. No public per-token or per-minute API price is stated
License
Not stated as a weight license. Site terms are linked from labs.riverside.com

Tags marked unsure are not stated on the lab's own pages.

Next steps

What would let us call it

Verdicts use only independent boards we already track. Two consecutive snapshots are required before any crown.

  • Publicly announced with a source
  • Usable by the public (API or weights)
  • Listed on an independent board we track
  • Enough results for a rank in a category
  • Holds across two snapshots (crowns only)
Self-reported

What the lab claims

Self-reported. Shown for context; not used in any verdict.

Lab claim: self-reported word-level control of pitch, tone, duration, pause, energy and pronunciation on both existing recordings and text-to-speech, while keeping the speaker's voice and room sound (Riverside Labs product page).

Receipts

Sources

LABTonality (Riverside Labs)labs.riverside.com/tonality · Nov 12, 2026LABRiverside Labslabs.riverside.comLABNadav Keyson on X: Tonalityx.com/NadavKeyson/status/2120938261875147169 · Nov 12, 2026LABNadav Keyson on X: Tonality APIx.com/NadavKeyson/status/2120938274642337848 · Nov 12, 2026

Tracking ends Feb 10, 2027 unless the API or weights become available.