murtaza-nasir/speakr v0.6.5-alpha

v0.6.5 - Model Flexibility

on GitHub

What's New

Separate Chat Model Configuration (#143)

• Configure a dedicated AI model for real-time chat interactions
• Use different service tiers: premium for chat, budget for summarization
• New environment variables: CHAT_MODEL_API_KEY, CHAT_MODEL_BASE_URL, CHAT_MODEL_NAME
• Separate GPT-5 settings for chat: CHAT_GPT5_REASONING_EFFORT, CHAT_GPT5_VERBOSITY
• Falls back to TEXT_MODEL_* settings if not configured

Custom Datetime Picker

• New themed calendar modal for date/time selection
• Consistent styling with application theme

Bug Fixes

• Fixed audio chunking not working after refactor (#140)
• Fixed username display showing "None" when full name not set (#138)

Upgrade

docker pull learnedmachine/speakr:latest
docker compose down && docker compose up -d

Compatibility

• ✅ Fully backward compatible with v0.6.x
• ✅ No configuration changes required
• ✅ New chat model settings are optional

Full Changelog: v0.6.4...v0.6.5