AUDIO NARRATION
30 VOICES, 80+ LANGUAGES

Audio narration is the simplest accessibility upgrade your blog can ship. Screen readers handle the basic case, but a real human-sounding narrator opens your content to a completely different audience: people who would rather listen during a commute, a workout, or while cooking dinner. Claude Blog wraps Google Gemini TTS so you can generate that narration directly from the same markdown that powers your post.

Pick a voice, run /blog audio generate, get a 24kHz mono WAV back. No re-recording, no DAW, no studio booking. The generator handles multi-speaker dialogue automatically when your post contains speaker tags, which makes interview-format and debate-format posts work without extra effort.

Commands

Voice selection

Voices map to tones. Charon is authoritative and works well for analytical or technical posts. Kore is warm and suits storytelling, lifestyle, or interview content. Puck is energetic and works for upbeat marketing pieces, product launches, and high-energy explainers. Aoede is breezy and reads well for casual narratives. Fenrir is deeper and grounded, useful for serious analysis. Each voice carries the same multilingual range, so pick the timbre that matches your post, not the language.

Costs and limits

Gemini TTS is free at the Google AI Studio level for normal blog cadence. Rate limits apply (requests per minute, characters per day), which is fine for a typical publishing rhythm of 1 to 3 posts per week. For podcast-scale workflows or full-archive backfills, the paid Gemini API tier is billed per million characters of input text at premium pricing. Output audio itself is not metered, only the input text is counted.