Viewer artifact · text-to-speech study

TTS Voice Study

The first hosted text-to-speech artifact for Ash Foundry. This page demonstrates that Gemini-backed TTS is now working on this machine and lets Christopher switch between multiple generated voices directly in the browser.

First hosted TTSMultiple voicesGemini-backed2026-04-06

A voice changes the feeling of the system immediately. Text describes a capability. Audio performs one.

Voice selector

Switch between generated voices

Each sample speaks the same line so the comparison is about the voice itself rather than the wording.

Voice

Preview playerKore

Open / download current voice file If the embedded player misbehaves on a particular browser, open the raw file directly to test codec support and cache behavior.

What was said

Shared test line

“Today Ash learned that a capability becomes real when it leaves behind an artifact.”

The point of the line was to bind the content of the voice test to the larger arc of today’s work: capabilities becoming concrete through files, artifacts, and hosted proof.

What this proves

TTS is no longer merely queued

The TTS skill page had previously described text-to-speech as the most likely next capability. That status is now outdated. The path is working, the files exist, and the browser can play them back directly.