BreezeVoice
Text to speech + subtitles · Windows
BreezeVoice is a text-to-speech app for Windows: type your text, pick a service and voice, and export audio with SRT subtitles. Edge, Azure, Alibaba Cloud Bailian, ElevenLabs and Windows system voices sit in one window with the same speed, pitch and volume controls, so you don't jump between dashboards.
Interface mockup; the shipping app may differ
Five voice services, one window
Edge and system voices need no account. Azure, Alibaba Cloud Bailian and ElevenLabs use your own API keys, billed by each provider. Keys are kept in Windows Credential Manager, never written to a file.
Edge Online
No key needed; sentence-accurate subtitle timing (unofficial Microsoft endpoint)
Azure Speech
Official Microsoft service with a free tier; pronunciation fixes via SSML phonemes
Alibaba Cloud Bailian
CosyVoice Chinese voices with adjustable speed, pitch and volume
ElevenLabs
Emotion tags, plus any cloned voices already in your account
Windows system voices
Works offline; basic quality, handy as a fallback
Features
Pauses and pronunciation fixes
Insert pauses of a set length that behave the same on every service, with subtitle timing shifted to match. Set the reading of Chinese characters with more than one pronunciation by hand.
Long text, split automatically
Paste a whole article. It is split at sentence boundaries, synthesized piece by piece and joined into one file, with a progress bar and cancel at any time.
Batch generation
Split by line, by blank line or by file, one audio file per item. A failed item does not stop the rest and can be retried on its own.
Multi-voice dialogue
Assign each line to a role, each role with its own service and voice. Lines are synthesized with gaps between them and exported as one audio file plus subtitles.
Tone director
Describe the delivery in one sentence, such as “start calm and slow, then speed up and finish angry”, and an LLM splits the script and marks speed and emotion for each part. Works with Alibaba Cloud Bailian, DeepSeek, Kimi, OpenAI or a custom endpoint.
MP3 / WAV + SRT subtitles
Export MP3 or WAV with an SRT file alongside. Edge uses sentence timings from the service for exact timing; other services are estimated by length, and each file is labelled exact or estimated.
Free and Pro: the free version covers single-text voiceover, up to 10 generations a day. Activating Pro unlocks batch generation and multi-voice dialogue with no daily limit.
Screenshots
Actual screens from the Windows app (Chinese UI).
Tone director
Describe the delivery in one sentence; an LLM splits the script and marks speed and emotion. Synthesize in one pass for a smooth flow, or per segment for clear contrast.
ElevenLabs voice library
Thousands of shared voices, filtered by gender, age, language, use case and keyword. Preview a sample, then use it in one click.
Alibaba Cloud Bailian voices
Close to 200 CosyVoice v2 / v3-flash / v3-plus voices, filtered by model, language and use case, with previews.
API keys and settings
All provider keys in one place with sign-up links, plus a monthly ElevenLabs quota check. Keys live only in Windows Credential Manager.
Listen to a sample
Generated by BreezeVoice (Mandarin).
“Need a voiceover? Hi, this is my voice! You are welcome to use it for your videos.” (spoken in Mandarin)