nuvocode/speaksy
A local-first, open-source speaking practice app.
Image
Networking
Machine learning & AI
0
424
nuvocode/speaksy repository overview
Speaksy
A local-first, open-source speaking practice app. Practice fluency by talking to an AI partner with your microphone and hearing spoken responses back.
Screenshots
Provider Configuration
Create a .env file to configure your AI provider:
# Local - Ollama (recommended)
AI_PROVIDER=ollama
OLLAMA_MODEL=llama3.2
# Local - LM Studio
AI_PROVIDER=lmstudio
LMSTUDIO_MODEL=local-model
# Cloud - Gemini
AI_PROVIDER=gemini
GEMINI_API_KEY=your_key_here
# Cloud - OpenAI
AI_PROVIDER=openai
OPENAI_API_KEY=your_key_here
# Cloud - Anthropic
AI_PROVIDER=anthropic
ANTHROPIC_API_KEY=your_key_here
# Cloud - Groq
AI_PROVIDER=groq
GROQ_API_KEY=your_key_here
Environment Variables
| Variable | Default | Description |
|---|---|---|
AI_PROVIDER | ollama | Provider: ollama, lmstudio, gemini, openai, anthropic, groq |
OLLAMA_MODEL | llama3.2 | Ollama model name |
LMSTUDIO_MODEL | local-model | LM Studio model identifier |
STT_PROVIDER | webspeech | Browser STT or whisper |
KOKORO_URL | http://localhost:8880 | Kokoro TTS endpoint |
KOKORO_VOICE | af_heart | Default voice |
Tag summary
Content type
Image
Digest
sha256:3aef82a74…
Size
59.3 MB
Last updated
4 months ago
docker pull nuvocode/speaksy