Need TTS API with SSML and custom pronunciation rules

Yes — you’ll want a TTS API that supports:

• SSML for prosody, pauses, emphasis, voice/style control
• Custom pronunciation via lexicons, phoneme tags, or text replacement rules

Good options:

1. Amazon Polly

   • Supports SSML
   • Supports lexicons (custom pronunciation dictionaries)
   • Good for production use and many languages

2. Google Cloud Text-to-Speech

   • Supports SSML
   • Supports custom pronunciation features in some workflows, though less flexible than Polly lexicons depending on use case

3. Microsoft Azure Neural TTS

   • Supports SSML
   • Supports custom lexicons/pronunciation and strong voice/style controls

4. ElevenLabs

   • Supports SSML-like control in some setups, but pronunciation customization is more limited/less standard than the cloud enterprise APIs above

If you need the best fit for SSML + custom pronunciation rules, I’d usually recommend:

• Amazon Polly if you want straightforward lexicon-based pronunciation control
• Azure TTS if you want strong SSML and enterprise speech control
• Google TTS if you’re already in GCP

Example SSML:

<speak>
  Hello, <break time="500ms"/> welcome to <emphasis level="strong">Acme Corp</emphasis>.
</speak>

Example custom pronunciation approaches:

• Lexicon file for words like product names or acronyms
• Phoneme tags for exact pronunciation
• Replacement rules in your app before sending text to TTS

If you want, I can give you:

• a comparison table
• a Python/Node example
• or the best API choice for your stack and budget