Podcast Generator Guide: Create a Studio-Quality AI Podcast From Any Topic
# Podcast Generator Guide: Create a Studio-Quality AI Podcast From Any Topic
Publishing a podcast used to mean booking a studio, coordinating guests, and spending hours in post-production. Vincony's Podcast Generator collapses all of that into a single form. Type a topic, hit generate, and within about a minute you have a two-host conversation — complete with distinct AI voices, a full transcript, and export options — ready to share.
This guide walks through every control in the tool, explains the pipeline behind the scenes, and gives you the tips that separate a generic episode from one that actually sounds researched.
How It Works
The generator runs four sequential stages, each visible as a progress step while your episode is being built.
Sources — If you supply URLs, Firecrawl scrapes the live content from each page and feeds it into the pipeline as raw research material. If you supply only a topic, the AI draws on its own knowledge and the latest available information.
Research — The scraped content (or the bare topic) is analyzed to extract key claims, facts, debate points, and background context worth covering in the episode.
Script — Gemini 3 Pro writes a natural back-and-forth dialogue between two named hosts: Alex (the interviewer, using the Roger voice) and Sam (the expert, using the Laura voice). The script length scales to the duration you chose — roughly 400 words for a two-minute episode, 1,000 for five minutes, and 2,000 for a full ten-minute episode.
Audio — ElevenLabs TTS renders each host's lines separately and returns the audio segments. The built-in player queues them in order so the conversation plays back seamlessly.
Each generation costs 10 credits. That same 10-credit cost applies each time you refine an episode, so plan your iterations before burning runs. Free-plan users get a limited monthly credit pool; Starter, Pro, Power, and Business plans each unlock progressively larger allowances — see pricing for the full breakdown.
Step-by-Step Walkthrough
Step 1 — Enter Your Topic
Open the Podcast Generator and type your topic in the Topic field (up to 500 characters). The more specific you are, the better the output. Compare these two:
- Too vague: "climate change"
- Much better: "How carbon capture startups are competing with reforestation in 2026"
A character counter appears at the bottom right of the field so you know how much room you have. If you want to jump-start an idea, click any of the sample topic chips — The future of AI in 2026, Bitcoin vs Ethereum, Remote work culture, or Space exploration milestones — to pre-fill the field instantly.
You can also trigger generation from the keyboard with Ctrl+Enter (or Cmd+Enter on Mac) once a topic is entered.
Step 2 — Add Source URLs (Optional)
For data-driven or niche episodes, paste up to five URLs into the Source URLs field and press Enter (or click the plus button) to add each one. These can be news articles, research papers, blog posts, or any publicly accessible page. Firecrawl will scrape their live content and your hosts will discuss what they actually find there — not a hallucinated summary.
Added URLs appear as removable badges beneath the input. Click the X on any badge to drop a source before generating.
Supplying URLs is especially effective when: - The topic is very recent and outside the model's training window - You want the episode grounded in specific data (earnings reports, studies, policy documents) - You are creating a product deep-dive or competitive analysis
Step 3 — Pick an Episode Style
Choose one of four styles by clicking the card that fits your goal:
| Style | Best for |
|---|---|
| Casual Chat | Accessible introductions, lifestyle or pop-culture topics |
| Deep Dive | Technical subjects, analytical breakdowns, research topics |
| News Briefing | Current events, crisp factual summaries |
| Debate | Controversial topics where two genuine perspectives exist |
Debate is particularly powerful for topics like AI regulation, investment strategies, or policy questions — the script explicitly assigns opposing viewpoints to Alex and Sam, which makes the audio feel like a real interview rather than a rehearsed FAQ.
Your style preference is saved automatically between sessions, so you will not need to re-select it on your next visit.
Step 4 — Set the Duration
Three duration buttons control the episode length:
- ~2 min — quick briefing, roughly 400 words of dialogue
- ~5 min — standard episode, roughly 1,000 words (the default)
- ~10 min — deep episode, roughly 2,000 words
Longer durations produce richer content but take proportionally more time to render. The generation progress note reads "Typically 30–60s depending on episode length" — expect the upper end for ten-minute runs.
Step 5 — Generate
Click the Generate Podcast button (or press Ctrl+Enter). The four-step progress bar — Sources → Research → Script → Audio — animates in real time so you can follow along. An elapsed-seconds counter keeps you oriented during longer runs.
While generation runs, Vincony saves your topic as a draft. If you accidentally navigate away and return, a draft-recovery prompt will offer to restore your topic text.
Step 6 — Listen and Review
When generation completes, the Your Episode panel appears with the PodcastPlayer component. The player streams each speaker's audio segment in sequence. Below the player you will see a word count and estimated reading time for the full transcript.
If ElevenLabs audio is unavailable (for instance, when the integration is not configured on an account), the tool displays a notice and falls back to transcript-only mode — you still get the full script, just without audio playback.
Step 7 — Refine the Episode
Below the player is a Refine this episode panel. You can reshape the output without starting from scratch using one of four quick chips:
- Go deeper on this topic — adds more detail and supporting evidence
- Make it more casual — loosens the tone and reduces jargon
- Add counterarguments — pushes Sam and Alex to challenge each other more
- Focus on practical examples — grounds abstract points in real-world scenarios
Or type your own direction in the free-text field and submit with Ctrl+Enter. Each refinement costs another 10 credits and regenerates the full episode with your adjustment baked into the prompt.
Step 8 — Export and Save
The Export and Share button opens a small menu with three options:
- Copy transcript — copies the full plain-text script to your clipboard
- Share — invokes the browser's native share sheet (or falls back to clipboard copy)
- Download .txt — saves the transcript as a text file named podcast-transcript.txt
Click Save (the heart icon) to add the episode to your Saved Episodes list. Saved episodes persist in your account and can be copied or deleted at any time from the collapsible panel at the bottom of the page.
To try a different angle without losing the current result, click New Episode to reset the form, or Regenerate to rerun the same topic and settings with a fresh pass.
Pro Tips
Stack URLs with a tight topic. The best episodes combine a focused topic with two or three authoritative URLs. The topic gives direction; the URLs give substance. Broad topics with no URLs tend to produce generic small-talk; specific topics with strong sources produce episodes that feel genuinely researched.
Use Debate for SEO-friendly content. A Debate-style episode on a contested topic produces natural keyword variation — both sides of an argument — which works well when you plan to publish the transcript as a blog post alongside the audio.
Save before you refine. Clicking Save before you run a refinement preserves the previous version. Since each refinement costs 10 credits, keeping a snapshot you can copy from is a useful safety net.
Chain with other tools. Take the downloaded transcript into Supercharged Transcription to clean up any artifacts, or feed a finished podcast outline into Songs to generate a matching intro jingle for the episode.
Match style to platform. Casual Chat uploads well to YouTube Shorts-style snippets; Deep Dive transcripts perform on text-heavy platforms like Substack; News Briefing works for LinkedIn audio posts; Debate is made for Reddit AMAs repurposed as audio.
What to Expect on Each Plan
Free-plan users can test the tool with the limited monthly credit pool — enough for a few short episodes to evaluate quality. Starter unlocks more credits and is workable for personal projects. Pro and Power users will comfortably generate and iterate multiple episodes per day. Business accounts have the headroom for content teams producing at scale.
If you run out of credits mid-workflow, the generate button will surface a credit-depleted message rather than silently failing — check pricing to top up or upgrade before a planned session.
Closing Thoughts
Stay ahead in AI
Get our weekly AI insights — tips, model comparisons, and guides delivered to your inbox.
No spam, unsubscribe anytime.
The Podcast Generator is one of the higher-value tools in the Vincony OS suite precisely because it stacks three difficult capabilities — live research, scriptwriting, and voice synthesis — into a single credit spend. Once you have the workflow down, producing a polished five-minute episode takes less time than writing a decent tweet thread on the same subject.
Pair it with the Audio suite for custom voice experiments, or use the transcript output as raw material for long-form content elsewhere. The episode is yours to publish however you like.