Azure-only workspace
The page filters the voice list to Azure Speech voices only, so style testing is not mixed with Google or other providers.
Azure Speech on TTSForge
A focused workspace for Azure-only voices, mstts styles, and Azure-friendly SSML iteration.
Write inner SSML only, insert Azure-friendly tags from the toolbar, validate your markup, and generate downloadable audio without leaving the page.
TTSForge / Azure
Azure Speech SSML Editor
Convert Azure-friendly SSML into speech with support for mstts styles, silence control, pronunciation, and more expressive neural voice testing.
Quick guide
1. This workspace is SSML-only; the page adds the Azure <speak> root and selected <voice> wrapper automatically.
SSML editor
Write inner SSML only. The page adds Azure speak and voice wrappers automatically when generating.
967/2000
Azure SSML Toolbar
Insert snippets quickly or wrap the selected text. The page auto-adds the Azure speak and voice wrappers for you.
Within limitYou do not need to type the root <speak> or selected <voice> tags.
Limit exceeded by 967 chars! Resets at .
Preset templates
Insert Azure-friendly structures with one click.
⌄
Preset templates
Insert Azure-friendly structures with one click.
Supported tags
A short Azure-oriented cheat sheet so you can test mstts and standard SSML tags faster.
⌄
Supported tags
A short Azure-oriented cheat sheet so you can test mstts and standard SSML tags faster.
break
<break>Insert a pause in the spoken output.
<break time="500ms"/>mstts:silence
<mstts:silence>Control silence around sentence or punctuation boundaries with Azure-specific timing.
<mstts:silence type="Sentenceboundary" value="200ms"/>say-as
<say-as>Read numbers, dates, and characters with the right interpretation.
<say-as interpret-as="date" format="mdy">03/18/2026</say-as>sub
<sub>Replace spoken output with a friendlier alias.
<sub alias="Azure Speech">Azure TTS</sub>phoneme
<phoneme>Override pronunciation using IPA or another phoneme alphabet.
<phoneme alphabet="ipa" ph="təˈmeɪtoʊ">tomato</phoneme>prosody
<prosody>Adjust speaking rate, pitch, or volume for a section.
<prosody rate="-10%" pitch="+1st">Hello everyone</prosody>emphasis
<emphasis>Emphasize a keyword or phrase.
<emphasis level="strong">very important</emphasis>mstts:express-as
<mstts:express-as>Azure-specific tag for changing style, styledegree, and role in a section.
<mstts:express-as style="customerservice" role="YoungAdultFemale">How can I help you today?</mstts:express-as>p
<p>Group content into a paragraph.
<p>Paragraph content</p>s
<s>Group content into a sentence.
<s>Sentence content</s>lang
<lang>Temporarily switch the language for a section.
<lang xml:lang="en-US">hello world</lang>bookmark
<bookmark>Place a bookmark marker for synchronization events in the audio timeline.
<bookmark mark="scene-1"/>Voice settings
Show only Azure Speech voices and keep generation aligned with the main TTS workflow.
Select language
Output
Audio player, request info, and file download.
No audio yet. The player and download button will appear here after a successful generation.
Tutorial Video
Popular AI Tools
Core tools to generate voice, build subtitles, and convert content across common workflows.
Text to Speech
Convert plain text into natural AI voice.
Google TTS Forge
A dedicated SSML workspace for Google Cloud Text to Speech.
SRT to Speech
Generate voiceover synced with subtitle timestamps.
Video to SRT
Extract subtitle blocks from video/audio.
PDF to Speech
Turn PDF content into listenable speech.
OCR
Extract editable text from images and scans.
Speech to Text
Transcribe speech and generate AI summaries.
Related Workflows You Can Run Next
Working with documents? Use PDF to Speech or Document to Speech to convert files directly into audio.
For subtitle-based video workflow, go from Video to SRT to SRT to Speech for timeline-synced dubbing.
For multi-character scripts, use Multi-Voice TTS to assign different voices in one script.
Review, play and download your generated audios. Files will be deleted after 90 days. Download now!
Total: 0 items
| Order | Time | Voice | Text | Pitch | Status | Actions |
|---|---|---|---|---|---|---|
| No items | ||||||
No items
Quick workflow
Designed for both first-time users and Azure SSML-heavy editing.
- 1
Step 1: Compose inner SSML
Write directly, start from a preset, or use the toolbar for Azure-friendly tags.
- 2
Step 2: Choose an Azure voice
Pick the language and the exact Azure neural voice you want to test.
- 3
Step 3: Adjust pitch
Use pitch when needed and keep the rest of the playback customization in the output player.
- 4
Step 4: Validate and generate
The form adds Azure speak and voice wrappers automatically before creating audio.
Tip: start with mstts:express-as and change one variable at a time, such as style, styledegree, or prosody.
Why use the Azure TTS page?
A cleaner workflow for Azure Speech users who need more than the home page form.
Azure SSML wrapper built in
You write inner SSML only. The page automatically adds the Azure speak root, mstts namespace, xml:lang, and the selected voice wrapper.
Style and role testing
Quick presets and toolbar snippets make it faster to test mstts:express-as, role, sentence timing, and prosody variations.
Same output workflow
Quota, generation, audio playback, download, and share-link logic stay aligned with the main TTS workflow.
Frequently Asked Questions
Q: How is this Azure page different from the main Text-to-Speech page?
A: The main page is broader and text-first. This Azure page is narrower by design: it focuses on Azure-only voices, Azure SSML, mstts styles, and faster iteration for expressive speech workflows.
Q: Do I need to type the <speak> or <voice> tags myself?
A: No. You only write inner SSML content. The page automatically wraps it with the Azure speak root, mstts namespace, xml:lang, and the selected voice tag before sending the request.
Q: Which Azure-specific tags are emphasized here?
A: The page highlights Azure-friendly tags such as mstts:express-as and mstts:silence, alongside standard SSML tags like break, prosody, say-as, sub, and phoneme.
Q: Will every Azure voice support every style or role?
A: Not always. Azure styles, styledegree, and role support depend on the specific neural voice. This page helps you test quickly, but voice-level capability still depends on Azure Speech support.
Q: Can I use plain text instead of SSML here?
A: This workspace is intentionally SSML-first. If you only need quick plain-text generation, the main Text to Speech page is a better fit.
Q: Can I download and share the result?
A: Yes. After a successful generation, you can play the audio, download it, and copy a share link from the output panel.