v3.6: Multi-language & Custom Endpoint Support

What's New in v3.6

New Features

• Optional API key for custom endpoints – API key is only required for OpenAI. Local/custom TTS servers (like Alltalk, Chatterbox) can work without authentication.
• Multi-language support – 54 languages now available in HA Assist pipeline settings.

Bug Fixes

• Skip API validation for custom endpoints (fixes compatibility with Alltalk V2, Chatterbox, etc.)

Installation

Available in HACS default repository. Search for "OpenAI TTS" and download.

v3.6b3: Fix custom endpoint validation

Bug Fixes

• Skip API validation for custom endpoints (fixes #52) - Custom endpoints like Alltalk V2 may not support the model used for validation. Validation now only runs for the default OpenAI endpoint.

v3.6b2: Optional API key for custom endpoints

What's New

• Optional API key for custom endpoints - API key is now only required for OpenAI. Custom/local TTS servers can work without authentication.
• Multi-language support (from v3.6b1) - 54 supported languages for HA Assist pipeline.

v3.6b1: Multi-language support

What's New

Bug Fixes

• Multi-language support for HA Assist pipeline (fixes #50)
  • Added 54 supported languages including Czech, German, French, Spanish, etc.
  • Languages now appear in Home Assistant Assist pipeline settings
  • OpenAI TTS auto-detects language from text (no configuration needed)

Installation

This is a beta release. Install via HACS by adding as custom repository and selecting the beta branch.

v3.5: Streaming, volume fixes, diagnostics

What's New

Features

• TTS Streaming – Reduced latency with streaming support (HA 2025.7+)
• Reconfigure flow – Change API key and URL without recreating the entity
• Sub-entries support – Multiple TTS agents under one integration (HA 2025.7+)
• Diagnostics – Added diagnostics support for troubleshooting
• Duration metadata – Reliable MP3 duration embedding using mutagen

Bug Fixes

• Fixed volume restoration timing for chime/normalize mode
• Fixed chime settings not applied from entity configuration
• Fixed voice/model changes not taking effect (cache key issue)
• Fixed ffmpeg metadata error (switched to mutagen library)
• Fixed "expected str" error on empty instructions field

Improvements

• Improved volume restoration logic and timing
• Complete translations (Czech, German, Greek, English)
• Code quality improvements and type annotations
• HACS and Hassfest validation workflows

v3.5b1: Volume timing fixes, TTS cache improvements, diagnostics

Bug Fixes

Volume Restoration Timing

• Fixed volume being restored too early when using chime or normalize features
• Volume restoration now correctly waits for full playback duration
• Added platform-specific timing buffers (Sonos: 500ms, Cast: 700ms)

Chime Configuration

• Fixed chime settings not being applied from entity configuration
• Service now reads chime/normalize settings from TTS entity subentry config

Voice Change Bug

• Added default_options property to TTS entity
• Voice/model/speed changes in config now properly invalidate HA's TTS cache

New Features

Duration Metadata with Mutagen

• Added reliable MP3 metadata embedding using mutagen library
• Duration stored in ID3 TXXX:duration_ms tag

In-Memory Duration Cache

• Added MESSAGE_DURATIONS_KEY for fast duration lookups
• Shared cache across all TTS entities

Diagnostics Support

• New diagnostics.py for troubleshooting
• Exports config entry data (API key redacted)

Code Quality

• Fixed deprecated asyncio.get_event_loop() calls
• Fixed bare except: clauses
• Removed duplicate constant definitions
• Added type annotations

reconfigure support, entity migration and handling fixes

3.4b5

entity migration fixes, strings

Bug fixes and stability enhancements.

3.4b2

Update manifest.json

bug fixes, translation updates

3.4b1

bug fixes, translation fixes

3.4b0

⭐ New Features in v0.3.4 Beta

• Volume Restoration – Automatically restores speaker volumes to their original levels after TTS announcements.
• Media Pause/Resume – Pauses currently playing media during announcements and resumes afterward (works with Sonos speakers).
• Sonos Integration – Automatically groups and ungroups Sonos speakers for synchronized announcements.
• New openai_tts.say Service – New service parameters including volume control and pause/resume options.
• Precise Audio Duration Detection – Improved timing for TTS playback with better synchronization.
• Media Player Targeting – Target speakers using areas, devices, or entity IDs for more flexible automations.
• Performance Optimizations – Improved audio processing for faster TTS responses.

Full Changelog: v3.3...3.4b0