Designed for multilingual families.
Hearthsay is designed for multilingual families. Original recordings can be preserved in any language you choose to record.
What is supported in your language
Support depends on language, provider availability, recording quality, and consent settings. Original recordings are always preserved.
Full Memory Support
Record, store the original, add a manual transcript, get a source-backed AI answer, and a translation if available.
Transcription
Automatic transcription with original and translated text when available.
Voice Playback
Familiar or synthetic voice playback through a supported provider.
Experimental
Record and store memories. AI and voice playback are limited, and originals are always preserved.
Language roadmap
Where we are. Where we are going.
Tracking our progress from 12 languages today to 30 by December 2026.
Now — June 2026
Early Access
12 languages
August 2026
Summer 2026
18 languages
October 2026
Fall 2026
24 languages
December 2026
Holiday 2026
30 languages
Now — June 2026 · 12 languages
Full UI, recording, transcription, and memory-grounded answers in all 12 languages. VoiceDNA and translation previews available.
August 2026 · 18 languages
Adding 6 languages spoken by major diaspora communities — full UI translation plus recording support for South Asian and African families.
October 2026 · 24 languages
Expanding into Southeast Asia, Eastern Europe, and the Middle East with voice playback quality review for each new locale.
December 2026 · 30 languages
Completing the 2026 roadmap with African and South/Southeast Asian languages — targeting families with fewer existing digital memory tools.
English
English- Provider
- ElevenLabs, Azure Neural, OpenAI
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
Español
Spanish- Provider
- ElevenLabs
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
हिन्दी
Hindi- Provider
- ElevenLabs
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
العربية
Arabic- Provider
- ElevenLabs
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
తెలుగు
Telugu- Provider
- ElevenLabs, OpenAI
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Not yet
中文
Chinese- Provider
- Azure Neural
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
தமிழ்
Tamil- Provider
- Azure Neural
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
日本語
Japanese- Provider
- Azure Neural
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
Français
French- Provider
- Google TTS
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
Português
Portuguese- Provider
- Google TTS
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
Русский
Russian- Provider
- Google TTS
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Available
বাংলা
Bengali- Provider
- Google TTS
- Original recordings
- Always preserved
- Translation
- Available
- Familiar voice
- Not yet
Language availability depends on the AI/voice provider you connect. Support levels are estimates and can change. Original recordings are always preserved.
Languages and voice
- Hearthsay is designed for multilingual families. Original recordings can be preserved in any language you choose to record. Automatic transcription, translation, and familiar voice playback depend on language support, provider availability, recording quality, and consent settings.
- If familiar voice playback is not available for a language, Hearthsay will never fake certainty. It will preserve the original recordings, show transcripts and translations when available, and clearly label any fallback voice.
- Hearthsay always preserves the original recording. AI voice playback depends on language, accent, provider support, consent, and recording quality.
Start with one story. One day, it may mean everything.
Record a single memory today and begin a private family vault — grounded, consent-based, and kept close.