Hey HN! I'm excited to share a preview of Podsynth (https://www.podsynth.com/). It's an open source canvas for testing out mixes of language and audio generation models. With it, you can test your llm/audio ideas before building a complicated app.
The initial preview features an "RSS → Audio Summary" workflow. It grabs recent items from an RSS feed, invokes a language model to create a summary, and generates TTS using an ElevenLabs text-to-speech model.
It's still very early, but I'm thinking to support:
- API endpoints to invoke your creations
- More models and more providers
- n8n, trigger.dev, or other scheduling/workflow integrations
Try it out, and let me know what you think!