How to write a task for a 'Role-based Podcast': fields, roles and duration
The 'Role-based Podcast' tool creates a podcast dialogue with host roles, segments and natural transitions. To get a useful result, it is important to formulate the task accurately. In this article we will look at what to write in the fields 'Your text or idea', 'Role or characters' and 'Duration or length', give examples and explain how to check the result and what limitations exist.
Open tool: Podcast script writer ↗Field 'Your text or idea': context and purpose
In the field 'Your text or idea' (up to 8000 characters) describe the topic, purpose and key points. The more specific, the better. Indicate who the podcast is for, what problem it solves and what the listener should learn. For example, for a podcast about beginner gardening you can write: 'Explain to beginners how to choose the first plants for a balcony, what mistakes are most often made and how to avoid them'. This sets the direction.
Avoid generic phrases like 'make it interesting'. Instead, list the topics to be covered and facts that cannot be distorted. If there are time or style constraints, indicate them here. Remember: the tool does not generate audio, only text dialogue, so it is important to describe the content of the conversation specifically.
Bad: 'Podcast about gardening'. Good: 'Five-minute dialogue of two hosts for beginners: which 3 plants are easiest to grow on a balcony, why they are suitable, and how not to overwater them'.
Field 'Role or characters': who speaks and how
In the field 'Role or characters' (up to 1000 characters) describe the hosts: their names, characters, manner of speech. This will help the tool create natural replies. For example: 'Anna - experienced gardener, speaks calmly and confidently; Igor - beginner, asks naive questions and sometimes jokes'. The clearer the roles, the less confusion in the dialogue.
If the podcast has guests or experts, indicate their role and style. You do not need to invent complex biographies - 2-3 characteristics are enough. The tool does not clone voices or create audio, so role descriptions affect only the text.
'Host Maria - practicing doctor, speaks clearly and without panic; co-host Pavel - journalist, asks clarifying questions. Friendly tone, no medical jargon'.
Field 'Duration or length': how long the podcast should be
In the field 'Duration or length' indicate approximate time in minutes or number of words. The tool does not generate audio, so duration affects text volume. For example, '5 minutes' approximately corresponds to 600-800 words of dialogue. If you need it shorter, set 2-3 minutes.
Note that the tool does not automatically check the chronometry, so after generation you will have to assess the length manually. If the text turns out too long, you can shorten it manually or reformulate the task. Do not specify unrealistic values, for example '1 minute' for a complex topic with three segments.
'Duration: 5 minutes. Need to fit into 700-800 words, so that when voiced you get approximately 5 minutes'.
Example of a bad and good task
A bad task often contains only the topic and gives the tool no guidance. For example: 'Make a podcast about healthy eating'. As a result, it may turn out a general dialogue without structure and specificity. A good task includes context, roles, duration and key questions.
Compare: 'Create a dialogue of two hosts for a podcast about healthy eating. Host Olga - a dietitian, speaks calmly; host Dmitry - a skeptic, asks questions. Duration 4 minutes. Discuss 3 myths about breakfast and give practical advice. No personalized recommendations'. Such a task will give structured and useful text.
Bad: 'Podcast about finance'. Good: '6-minute dialogue: host-economist and host-student discuss how to manage a budget for a month. Roles: economist explains simply, student asks everyday questions. Goal - educational, without specific investment advice'.
How to check the result
After generation read the dialogue and check if it matches the task. Pay attention to roles: do the characters speak in accordance with the description? Are there segments and natural transitions? If something is wrong, adjust the fields and generate again. The tool does not save history, so copy successful text immediately.
Check factual accuracy: are there any fictional data or dangerous advice. In finance, health and law topics the text should remain educational and not give personalized recommendations. If the dialogue contains such advice, edit manually. Remember that the model may make mistakes, and queue and availability may change.
After generation you see that the host-beginner speaks like an expert. Fix the role description: 'Igor - beginner, asks simple questions' and repeat generation.
Tool limitations
The tool works only with text: it does not create audio or video, does not clone voices and does not generate photorealistic images. Vector images are available only in SVG and PNG formats. Speech can be synthesized via eSpeak to WAV, and subtitles - in SRT. No DOCX or PDF export, batch processing, automatic saving and publishing.
Maximum length of the first field - 8000 characters, detail fields - 1000 characters. For speech - up to 3000 characters, for images - up to 4 MB, for browser audio - 20 MB or 5 minutes. The tool does not require registration and has no premium plan. Model responses may contain errors, so always check the text.
You cannot upload a document for analysis or get a ready-made audio file of the podcast. You can only download text in TXT.
An accurate task for the 'Role-based Podcast' saves time and gives a quality dialogue. Fill in the fields specifically: describe the topic and goal, roles and characters, duration. Check the result for compliance and factual accuracy. Remember the limitations: the tool creates only text, not audio. Use examples from the article as a template, but adapt them to your topic. This way you will get a useful podcast script with segments and natural transitions.
Open tool: Podcast script writer ↗Interface translations are machine-assisted and may need refinement.