I built GeminiTTS, a simple web tool for generating natural speech with Gemini TTS

9 月 30 日
 qqqvexxdh

I recently built GeminiTTS, a small web tool for turning text into natural-sounding speech with Gemini TTS.

Try GeminiTTS online

I built it because I wanted a simple way to use Gemini's text-to-speech models without setting up an API, writing code, or dealing with audio generation scripts every time.

With GeminiTTS, you paste in your text, choose a voice, adjust how you want it to sound, and generate the audio directly in the browser. You can describe the speaking style in natural language, such as calm, energetic, professional, conversational, or storytelling.

It also supports multi-speaker audio, so you can assign different voices to different speakers and generate conversations instead of creating each voice separately.

Some practical use cases:

What I like most about Gemini TTS is the controllability. Instead of only choosing a voice and changing speed or pitch, you can describe how the line should be spoken. This makes it much easier to experiment with tone, pacing, emotion, and conversational styles.

TTS is still not perfect. Long scripts, unusual pronunciations, or very specific speaking styles sometimes need another generation or a slightly different prompt. Breaking a long script into smaller sections usually gives me more predictable results.

This is an independent project and I am still improving the generation workflow, voice controls, and long-form audio experience.

If you use text-to-speech for videos, podcasts, apps, or AI projects, give it a try. Feedback and difficult test cases are welcome.

592 次点击
所在节点    llms.txt
0 条回复

这是一个专为移动设备优化的页面(即为了让你能够在 Google 搜索结果里秒开这个页面),如果你希望参与 V2EX 社区的讨论,你可以继续到 V2EX 上打开本讨论主题的完整版本。

https://www.v2ex.com/t/1245791

V2EX 是创意工作者们的社区,是一个分享自己正在做的有趣事物、交流想法,可以遇见新朋友甚至新机会的地方。

V2EX is a community of developers, designers and creative people.

© 2021 V2EX