I haven’t been able to work on my project for over two weeks, which requires generating audio with a British accent. Until now, I’ve only been using Gemini 2.5 Pro TTS, which I consider to be the best model currently available. This is the sole reason I pay for the Gemini AI Pro Plan.
The main problems I’ve encountered are micro-cuts, abrupt short interruptions, unnatural transitions, and a lack of consistency in the generated audio.
I’ve seen that other users have reported this problem, but so far, I haven’t seen a fix.
I’m aware that a new model, Gemini 3.8 TTS, has been released. The problem is that while this model doesn’t have these issues, it presents two problems for my project. First, it doesn’t offer a British voice option. The speaker profile can only be configured with something like “British accent” or “London accent.” Even then, the output is very inconsistent, and no matter how you do it, it will always produce a mix of an American and a British accent.
However, the main problem with this model is that after about 20 seconds, a very loud whistle and a metallic sound begin to appear, with significantly degraded audio quality. Right now, I think that while it’s a good model for adding more life and naturalness to speakers, it doesn’t even reach the quality level of the Gemini 2.5 Pro TTS.