How to Turn Your Podcast into Short Videos — A Complete Guide

Converting podcast audio into short videos takes three steps: clip selection → visual generation → FFmpeg composition. This guide covers the full pipeline with real specs, credit costs, and a decision table pointing you to the right deep-dive article for your specific needs.

TL;DR

• Input: MP3/WAV/M4A (≤200 MB and up to 90 minutes) or a podcast RSS feed URL
• AI surfaces up to 5 clip candidates with viral scores
• 4 visual styles: sketchnote / cinematic / flat_graphic / manga
• Output: 1080×1920 9:16 MP4, subtitle position Top/Bottom selectable
• Free plan: 50 credits/month — enough to test the full workflow end-to-end

Why Podcasts Need Short Video

Audio content is deep, but discoverability is hard. Short video algorithms actively surface content to non-followers, making vertical clips an efficient acquisition channel for new listeners. The key advantage for podcasters: you already have the raw material.

Repurpose existing episodes — no filming or scripting required

Multiple posts from one audio file — up to 5 clip candidates per episode

One MP4, three platforms — YouTube Shorts, TikTok, Instagram Reels

Which Article Should You Read? — Decision Table

Step 1: Upload Your Audio

Accepted input:

• File: MP3 / WAV / M4A — maximum 200 MB and 90 minutes
• RSS: paste a feed URL → get up to 100 episodes listed (episodes over 90 minutes are auto-excluded)

For RSS imports, select the episodes you want to process individually or use batch selection.

Step 2: STT Transcription + Clip Analysis — 5 Credits

After upload, Google Cloud Speech-to-Text v2 (model: chirp, word-level timestamps) transcribes the full audio. Then Gemini 3 Flash analyzes the transcript and returns up to 5 clip candidates, each containing:

Step 3: Visual Generation — 5 or 15 Credits

Gemini 3 Flash Image generates still images based on each clip's text content. Choose from 4 styles:

Generation modes:

Standard (1 frame) — 5 credits: A single image per clip

HQ Storyboard (3 frames) — 15 credits: Hook → Evidence → Payoff across three frames, each with an auto-composed headline overlay. Significantly better at stopping the scroll and driving saves.

Step 4: Video Generation — 0 Credits

FFmpeg composites the final video to these specs:

Subtitle position Top keeps captions clear of platform UI elements that appear at the bottom of the screen; Bottom is the standard placement.

Bonus: Article Generation — 3 Credits

From the same transcript, Faceless.fm can generate long-form posts for X / note / LinkedIn. Run video and article generation from a single episode and ship content across every channel at once.

Credit Reference

Who This Is For

Podcasters: Convert existing episodes directly to shorts — no new recording required

Creators who dislike filming: No camera, lighting, or video editing software needed

High-frequency posters: Up to 5 clips per episode means multiple posts per week from one audio file

RSS-heavy publishers: Batch-import from a feed URL and process multiple episodes at scale

Multi-channel publishers: Generate video and written posts from the same transcript in a single session

When It Might Not Be the Right Fit

Live or breaking-news content: STT → AI analysis → generation takes time; real-time reaction content does not fit this pipeline

Brand campaigns with strict visual guidelines: The 4 built-in AI styles may not match a specific brand system; fully custom design needs a different tool

Poor-quality source audio: Low-quality recordings reduce STT accuracy, which degrades clip candidate and hook-text quality downstream

Episodes over 90 minutes: Split the recording before import so each source stays within the processing window

Related Articles

Turn an Audio-Only Podcast into Short Videos (No Footage Needed)

Automate Your Podcast RSS into Shorts and Articles

What Are Faceless Videos?

Try It on Faceless.fm

https://faceless-fm.com — Free plan includes 50 credits/month. Upload an MP3 and see what clip candidates the AI surfaces in minutes. Upgrade to Pro (¥2,980/month) for 300 credits and batch RSS processing.