Languages and voices
WebVoice groups voices by language so you can pick the locale that matches your content — from widely used languages to several regional accents and gender options where the underlying engine exposes them. You are not limited to a single default voice: switch per project or per sentence when experimenting in the studio.
The exact catalogue depends on the active TTS backend (for example high-quality neural runtimes with ONNX or cloud providers). The dashboard shows only voices that are available for your account, with readable names so you can compare timbre and style.
Speed, clarity and workflow
You can adjust speech rate within a safe range so announcements sound natural, tutorials stay understandable, and short prompts remain punchy. This is useful for accessibility, localization (some languages sound better slightly slower), and matching the pacing of a video or IVR menu.
MP3 output and download
Generated audio is produced as compact MP3 by default in many flows, which is ideal for sharing, embedding in presentations, or sending by email without huge file sizes. From your generation history you can download the files again later, so you do not need to regenerate the same clip if you still have credits logged against that item.
For developers, the REST API returns Base64-encoded audio in JSON responses, so you can decode to MP3 bytes on the server or client and save or stream as you prefer.
API and automation
The same TTS engine is exposed through authenticated API calls: send text, voice id, language and speed; receive duration and audio payload. That lets you batch newsletters, voice menus, or app notifications without using the web UI.
At a glance
- Broad language coverage and multiple voices per language where supported
- Adjustable speed for clarity and style
- MP3 downloads from history plus API-friendly encoding
- Credits scale with audio length — transparent rules in the pricing section