This workflow integrates the notdaniel/voxtral-small-24b-2507 model from Replicate to generate audio content from provided inputs. It handles API authentication, creates predictions, polls until completion, and outputs the final generated audio file.
Benefit: Securely connects your workflow to Replicate’s API and ensures only authorized requests are made.
🌐 Create Prediction → Sends a request to Replicate’s API with parameters like:
Benefit: Starts the audio generation process with configurable input and settings.
🆔 Extract Prediction ID → Captures the unique prediction ID and endpoint for polling.
⏱️ Wait → Pauses for 2 seconds before re-checking.
📡 Check Prediction Status → Polls Replicate’s API to see if the audio generation is done.
✅ Check If Complete →
Benefit: Efficiently manages asynchronous audio generation, ensuring the workflow only proceeds when results are ready.
📝 Process Result → Extracts and structures final output data:
Benefit: Provides a clean, structured output that can be used in follow-up automations (e.g., sending audio to users, storing in a database, or sharing via email).
| Section | Purpose | Key Nodes | Benefit |
|---|---|---|---|
| ⚡ Trigger & Authentication | Start workflow & authenticate | Manual Trigger, Set API Key | Secure execution |
| 🎛️ Create Prediction | Submit audio generation request | Create Prediction | Start model processing |
| ⏳ Polling & Status Tracking | Monitor prediction progress | Extract Prediction ID, Wait, Check Prediction Status, Check If Complete | Ensures reliable completion |
| 🎧 Process Result | Format and deliver output | Process Result | Clean audio result ready for use |
