Audio reference module
Inspect the audio task path, then adapt it to your SaaS.
Why this module is here
This is a working audio workflow for validating the shared SaaS foundation. It is not a separate finished business or proof of a live provider integration.
Try it locally
Follow the reference demo, select the audio mode, create a short task, inspect its status and actual output, and confirm the credit change. Local results are labelled simulations.
Configure a real provider
The audio adapter lives in app/server/ai.ts. Configure the supported model, API key and cost table, then test the expected input/output format, size limits and error states. Website, API and Skill all use the same task/credit system.
| Feature | Credits per task | Meter |
|---|---|---|
| Thoughtful writing | 5 | tokens |
| Fresh perspectives | 20 | images |
| Ideas with a voice | 10 | seconds |
| A conversation that continues | 5 | tokens |
| Edit a private image | 20 | images |
| Understand an image | 5 | tokens |
| Answers with sources | 5 | queries |
Adapt it to a product
Describe the customer task, real example, limitations, pricing and output retention. If requirements differ from the reference module, implement the corresponding logic and recovery. Customize the project.