Guide - Podcasts - Radio Manager

What is a podcast?

The Podcasts module generates complete audio episodes in which one to three voices converse, with jingles and background music. Ideal for a themed show, a debate or an interview produced entirely by AI.

This feature requires the "podcasts" permission. You will find it in the Production > Podcasts menu.

A two-stage process

Creating a podcast takes place in two successive steps: first the script (the episode text), then the audio (the sound production). You review and edit the text before committing to audio generation.

Good to know: The page refreshes automatically during processing: there is no need to reload it — progress is displayed on its own.

Configuring the episode

Fill in the main parameters:

"Title" (required, 255 characters maximum).
"Theme/Topic" (required): the text area describing the episode content.
"Style": Conversational, Informative, Debate or Interview.
"Tone": Casual, Serious or Humorous.
"Target duration": ~2, ~5 (default), ~10 or ~15 minutes.
"Additional instructions" (optional): specific directives for the AI.

Example

Studio Nova is launching a weekly duo show about music. Here is the configuration for the first episode.

Title: Nova Sessions - The vinyl comeback
Theme/Topic: "Why people under thirty are buying vinyl records again. Price, listening ritual, sound quality, collector appeal. Close with what this says about our relationship with music."
Style: Conversational
Tone: Casual
Target duration: ~10 minutes
Additional instructions: "no made-up sales figures; alternate speaking turns every three or four sentences"

Why: listing the theme as bullet points gives the AI a structure to follow, resulting in a well-organised episode. A vague prompt such as "talk about vinyl" produces ten minutes that go in circles — and you pay the same price either way.

Title, theme, style, tone, target duration and additional instructions Click to enlarge

Choosing the mode and voices

The "Mode" determines the number of speakers: Solo, Duo or Trio. Depending on your choice, one to three voice panels appear. Each panel includes:

a Voice;
an Intonation (optional);
a Role as free text, with default values: "Host", "Co-host", "Devil\'s advocate".

Example

The same episode, in Duo mode. Two voice panels appear, and you replace the default roles.

Voice 1 — a calm female voice, intonation Standard, role: Host.
Voice 2 — a livelier male voice, role: Record shop owner, passionate and slightly critical.

Why: the Role field is free text, and the AI uses it to write the lines. "Record shop owner, passionate and slightly critical" creates a genuine counterpoint; leaving "Co-host" as is produces two voices saying the same thing in turns.

The three modes: Solo, Duo and Trio Click to enlarge

One panel per voice: selection, intonation and role Click to enlarge

Sound branding

You can enhance the episode with:

an Intro jingle and an Outro jingle, selected from "My Files";
Background music (minimum 60 seconds);
a "Music volume" as a percentage (10% by default).

A summary permanently displays the "Estimated cost: ~X credits".

Example

The sound branding for "Nova Sessions", kept identical from one episode to the next.

"Nova Sessions" intro jingle→ Duo dialogue (~10 min)→ Outro jingle

Intro jingle: the Nova Sessions intro file, previously imported into "My Files".
Background music: a calm instrumental bed.
Music volume: 10% — the default value.

Why: intro and outro jingles are chosen from "My Files": prepare them once and they will be reused for every episode, giving the series a consistent identity. Keep the background music at 10%: any higher and it competes with the dialogue instead of supporting it.

Intro jingle, outro jingle, background music and its volume Click to enlarge

Generating the script then the audio

Click "Generate script": the AI writes the episode, presented segment by segment and fully editable. An "Improve" button lets you rework each segment individually.
Once you are satisfied with the script, click "Generate audio": production is carried out in batches, then the segments are assembled and mixed into a single episode.

Tip: Read through and improve each script segment before launching the audio: this is the step that costs no credits and makes all the difference to the final quality.

Example

The "Nova Sessions" script has been generated. Here is the review to carry out before clicking "Generate audio".

Generate script→ Review and "Improve" weak segments→ Generate audio

Opening segment: too long — you shorten it directly in the text area.
Segment 4: the record shop owner\'s comeback falls flat — you click "Improve" for that segment alone.
Closing segment: you manually add the teaser for next week\'s episode.

Why: the script step is free; the audio step is not. Correcting a segment after audio generation means paying again for the entire episode.

The summary, the estimated cost, then "Generate script" and "Generate audio" Click to enlarge

The "My Podcasts" list

The "My Podcasts" page opens with a two-tab panel — "Upload" to import an already-produced episode and "AI Creation" to generate one — followed by a filter bar: search, status ("All statuses", "Completed", "In progress", "Draft", "Error") and category, with "Filter" and "Reset" buttons.

The table displays, for each episode, the cover art thumbnail, the "Title", the "Mode" (Solo, Duo or Trio), the "Category", the "Voice(s)", the "Duration", the "Status" and the "Date". At the start of each row, two quick-listen buttons: "Listen with music" and "Listen voice only".

End-of-row actions: "Download ZIP" (both versions of the episode), "View details" (the episode page, with the script segment by segment), "Edit title and cover art" and "Delete".
Click the category on a row to change it or create a new one on the spot.
Select multiple rows (or the header checkbox) for bulk actions: "Category" (add, remove, replace the category, or assign the same cover art to the entire selection) and "Delete".

The cover art is the image displayed next to the title in the radio player when the episode is on air. Without cover art, the category image and then the station logo are used instead. Accepted formats: JPG, PNG or WEBP, maximum 5 MB. It can also be set on the episode page, in the "Cover art" block, using "Add an image" / "Change image" and "Remove".

Tip: Assign the same cover art to all episodes in a series (multi-select, then "Category" > assign cover art): in your website and app player, the series becomes instantly recognisable at a glance.

Credits

1 credit = 3 seconds of audio. For a podcast, the estimate corresponds to approximately 70% of the target duration (an episode includes silences and music). If the estimate exceeds your balance, generation is refused.

Creation requires an active subscription or a credit balance greater than zero; otherwise you will be redirected to the pricing page.

Example

You compare the four target durations before launching the Studio Nova series.

~2 minutes: approximately 28 credits.
~5 minutes (default): approximately 70 credits.
~10 minutes: approximately 140 credits.
~15 minutes: approximately 210 credits.

The "Nova Sessions" duo episode of 10 minutes therefore costs around 140 credits, regardless of the number of voices: it is the duration that matters, not the Solo, Duo or Trio mode.

Why: for a first episode, start with a ~2-minute version using the same voices and the same sound branding. You validate the result for 28 credits instead of 140, then relaunch at 10 minutes once the settings are right.