Step 01
Setup
Create a feed in BeaconBeam, then copy its RSS URL into your podcast app—Pocket Casts, Overcast, Apple Podcasts, or whatever you already use. Your show lives there like any other subscription.
AI-narrated podcast feeds
Generate a private podcast out of any text—get it in Pocket Casts, Overcast, or wherever you already listen to your podcasts.
MCP available
Hand your feed to your agents
Claude · ChatGPT · Perplexity + more
How it works
Step 01
Create a feed in BeaconBeam, then copy its RSS URL into your podcast app—Pocket Casts, Overcast, Apple Podcasts, or whatever you already use. Your show lives there like any other subscription.
Step 02
Paste or write your script and start generation. When narration finishes, the episode lands in your feed—open your podcast app and it’s in your pocket, ready to play.
Step 03
Connect an agent via MCP: create episodes from scripts, trigger generation, and check status without living in the UI—once your feed is set up, the whole pipeline can run programmatically.
Demo video
Use cases
Some examples of what is possible with agents + BeaconBeam.
Case 01
Longreads from read-list get batched into a feed so the queue actually clears.
Case 02
Weekly paper summaries compiled and narrated to effortlessly keep in touch with what’s new.
Case 03
CEO has weekly reports from all departments aggregated to a weekly summary.
Case 04
New issues of a newsletter automatically set up to be narrated by the agent as they’re published.
Case 05
Monthly highlight compilation from Kindle exports transformed to a private "reading log" show in Overcast.
Case 06
Course notes get narrated into a semester feed so semester revisions are easily available in transport or on the go.
Case 07
Daily journal entries are compiled and narrated at the end of the month for reflection and archiving.
Case 08
Monthly compilation of shipped features and fixes for everyone to keep in touch with what changed in the product.
Case 09
Agent tracks competitor launches, pricing changes, and blog posts, to narrate a short weekly “what moved in the market” episode.
Agent control
BeaconBeam speaks MCP—the protocol agents use to operate outside services. Connect Claude, ChatGPT, or your editor once, and the agent does the whole job: read the page, draft the script, publish the episode.
It does not have to stay a conversation. Ask the agent to repeat the job on a schedule—every Friday, or each time a new paper lands in your reading folder—and the feed fills itself while you are somewhere else.
> You
summarise this paper and put it in my Papers feed
attached · scaling-laws-for-small-models.pdf · 28 pages
> Agent
Read it. I’ll write a spoken summary and publish it with the BeaconBeam MCP tools.
check_episode · ok · 8,400 chars · 78,200 / 120,000 remaining
> Agent
Here is the draft — 8,400 characters of narration, about 10 minutes of audio. You have 78,200 of 120,000 included characters left this period, so this fits with room to spare. Publish it?
title · “Small models, long training — what the numbers show”
“Today’s paper tests a claim you hear a lot: that a small model, trained long enough on good data, catches up with a much larger one. The authors run twelve benchmarks at four model sizes. On reading comprehension and translation the gap nearly closes. On anything that needs counting it does not…”
> You
let’s do it
> Agent
Published. The audio is generating now and will show up in your podcast app shortly.
Narrating options
Play narration examples and choose the voice that suits you the best.
British · warm & clear
American · bright & conversational
British · crisp & authoritative
American · calm & steady
Pricing
$20 / month
Generate a private podcast out of any text—get it in Pocket Casts, Overcast, or wherever you already listen to your podcasts.