Disclosure: this article contains no affiliate links. All links are plain links, and we earn nothing if you click them. See our affiliate disclosure. Based on Google’s official AI Studio docs, pricing page and launch posts, checked November 14, 2026.
AI-assisted: AI helped gather and draft this research; the facts were checked against the cited sources.
Quick answer: On November 4, 2026 (US time; early November 5 in Malaysia), Google made Voice Remixing for Gemini 3.8 Flash TTS available in Google AI Studio and the Gemini API. You pick an existing voice and describe what to change, like the accent, pitch or pace. Using it through the API requires a paid-tier project. We haven’t yet been able to confirm whether it works in the free AI Studio web app.
What Google launched
Google posted on X that Voice Remixing “is now live in @GoogleAIStudio and the Gemini API.” You can start from a preset in its library of 2,000+ voices or from a voice you designed yourself. Then you prompt it to adjust timbre, pitch, pace and accent.
In AI Studio, Google says you try “Remix this voice.” You can then audition three generated takes side by side, pick your favorites to keep remixing, and save them to reuse across scripts and multi-speaker scenes.
The feature had been listed as “coming soon” when Google launched the 3.8 TTS models on September 23.
How it works
According to the Voice remixing docs, you give it a base voice and a text instruction describing what to change. The docs’ example asks for a voice that sounds “slightly older and more grounded, with a warm British accent and measured audiobook narration pacing.” It creates a new saved voice and leaves the original alone. You can use the saved voice with Gemini 3.8 Flash TTS or 3.8 Flash-Lite TTS, and you can remix it again.

What it costs
From Google’s pricing page for Gemini 3.8 Flash TTS:
| Free tier | Paid, through Dec 31, 2026 | Paid, from Jan 1, 2027 | |
|---|---|---|---|
| Text input (per 1M tokens) | Free | $0.50 | $1.00 |
| Audio output (per 1M tokens) | Free | $9.00 (about $0.00225 per 10s) | $18.00 (about $0.0045 per 10s) |
| Used to improve Google products | Yes | No | No |

Is it free in the web app?
The docs don’t say directly. Here’s what they do say:
- The Voice remixing docs state that remixing “requires a Paid Tier (Tier 1 or higher) Gemini API project.”
- The same page says you can remix voices interactively in Google AI Studio. It doesn’t say whether that also needs a paid tier.
- The pricing page says “Google AI Studio usage is free of charge in all available regions,” and lists the 3.8 Flash TTS free tier as “Free of charge” for both input and output.
We haven’t been able to confirm whether remixing works on a free AI Studio account without billing turned on.
The catches
- API remixing requires a paid tier, as covered above.
- There’s a cap of 200 saved voices per project. It covers designed, remixed and replicated voices together. A saved voice expires one year after you last use it (docs).
- On the free tier, your content is used to improve Google products (pricing).
- Voice cloning is a separate feature, and Google’s two pages differ on where it works. The voice replication docs say replication through the API needs a paid-tier project and is available in all supported Gemini API regions, including the EEA, the UK and Switzerland. Google’s September 23 launch post says voice replication through AI Studio is not available in Illinois, Texas, the EEA, the UK, Switzerland and India.
Who it helps
Voice Remixing is useful for faceless videos, audiobook narration and short ads. The model page lists Standard Malay and Chinese (Simplified and Traditional) among its 130+ languages. Rather use a no-code tool than AI Studio? See our Fliki review for an AI video maker with built-in voiceovers.
FAQ
Is Voice Remixing free? The TTS model has a free tier. Remixing through the API requires a paid-tier project. We haven’t verified whether it works in the free web app.
What will it cost from January 1, 2027? $1.00 per 1M text input tokens and $18.00 per 1M audio output tokens.
Can it speak Malay or Chinese? Yes. Both are on the model’s language list.
Is the audio watermarked? Google says all audio from Gemini Audio models carries a SynthID watermark.
Sources
- Google on X, launch post (Nov 4, 2026): https://x.com/Google/status/2118027058794148099 (checked 2026-11-14)
- Google on X, AI Studio steps (Nov 4, 2026): https://x.com/Google/status/2118027061948547270 (checked 2026-11-14)
- Google AI Studio docs, Voice remixing: https://aistudio.google.com/docs/voice-remixing (checked 2026-11-14)
- Google AI Studio docs, Voice replication: https://aistudio.google.com/docs/voice-replication (checked 2026-11-14)
- Gemini Developer API pricing: https://aistudio.google.com/docs/pricing (checked 2026-11-14)
- Gemini 3.8 Flash TTS model page: https://aistudio.google.com/docs/models/gemini-3.8-flash-tts (checked 2026-11-14)
- Google blog, “Gemini 3.8 text-to-speech says hello” (Sep 23, 2026): https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-text-to-speech/ (checked 2026-11-14)
Related: Fliki review · free AI meeting notes with Fireflies · Fireflies Talk comes to mobile
Last checked: November 14, 2026 (MYT).
2 thoughts on “Gemini Voice Remixing Is Live: Tweak AI Voices With a Prompt”