On 22 September, Google made Gemini 3.8 Flash TTS and Flash-Lite TTS generally available. TTS means text-to-speech: you provide a script and receive an audio file.
The larger model supports 130 languages, including Swedish. It can read with one or several voices and follow descriptions of style and pace. That makes it possible to create early versions of audio guides, stories and dialogue before anyone records them for real.
This is mainly a tool for creators and developers. Google offers it through the Gemini API and Google AI Studio, not as a new button in the regular Gemini chat. A free tier is available, but limits apply and paid use costs money. According to the pricing information, content on the free tier may be used to improve Google’s products.
The models can also replicate a voice, but Google requires consent verification. That is an important boundary. Synthetic speech should also be labelled clearly when a listener might otherwise believe that a real person said it.