EchoScript : app.py: sources=[upload, microphone] added to the gr.Audio component. Gradio renders this as a tabbed widget: one tab for file upload (drag-and-drop or browse), one tab with a record button for the microphone. Added format=wav so microphone recordings are always saved as wav regardless of browser - this ensures a consistent, known extension that passes through validate_extension without any additional handling.No changes needed to services/audio.py - .wav was already in SUPPORTED_EXTENSIONS from day one, so microphone recordings go through exactly the same validation and transcription path as uploaded files. The time-window fields also work with microphone recordings, so you can record a long clip and transcribe only a specific portion if needed.

app.py

@@ -357,7 +357,7 @@ with gr.Blocks(title="EchoScript") as demo:
 **Upload Audio → Select Audio Window → Detect Language & Generate Transcript
 → Preview & Choose Languages → Generate Translations → Copy / Download**
 
-<sub>build: 2026-07-02 01:21 UTC &middot; fixed fa/pt/tr model names &middot; short labels &middot; Select All</sub>
+<sub>build: 2026-07-02 23:46 UTC &middot; microphone input added</sub>
         """
     )
 
@@ -365,11 +365,12 @@ with gr.Blocks(title="EchoScript") as demo:
         with gr.Column(scale=1):
             gr.Markdown("### Upload Audio")
             audio_input = gr.Audio(
-                label="Drop audio file here or click to browse",
-                sources=["upload"],
+                label="Upload a file or record from microphone",
+                sources=["upload", "microphone"],
                 type="filepath",
+                format="wav",
             )
-            gr.Markdown("Supported: mp3 &middot; wav &middot; m4a &middot; flac")
+            gr.Markdown("Upload: mp3 &middot; wav &middot; m4a &middot; flac &nbsp;|&nbsp; Microphone: recorded as wav")
 
             gr.Markdown("### Processing Window")
             with gr.Row():