Product Updates

You can now email someone a link to create a voice clone directly in your account, without them ever signing up for Jellypod.

Say you're producing an episode that needs your colleague as a subject matter expert. Invite them to record their voice as a Jellypod host, they upload a short sample from their phone, and a new host with their voice shows up in your library, ready to drop into the episode.

How to use it

1. On the Hosts page, click Invite a guest and enter the guest's name and a short backstory.
2. Copy the invite link, or click Send email to deliver it from Jellypod.
3. Your guest opens the link, no Jellypod account required, and uploads or records voice samples directly.
4. When they submit, the new host lands on your Hosts page and you get an email letting you know it's ready.

Manage pending invites at Hosts → Invites: resend the email, copy the link again, or revoke it. Links are single-use and expire after 24 hours.

You can now import an externally-generated script, letting Jellypod match script speakers to your account's existing host voices. If a speaker cannot be matched to an existing host, Jellypod will intelligently create a new host with a suitable voice from our Voice Library.

Say you wrote a multi-host interview in ChatGPT or a Google Doc. Import the script to Jellypod and as long as you put an existing host name on each speaker's lines, Jellypod will match it correctly. Any names where Jellypod can't find a match and only exist in the script are created automatically with a fitting voice and a short backstory. No setup beforehand required.

How to use it

1. Open the Create Something New page and select Upload Existing Script.
2. Add your script. Paste it in, or upload a plain text, Markdown, or PDF file. Marking speakers in brackets like [Alex] gives the cleanest results, but plain prose works too - Jellypod detects who is speaking.
3. Review the preview. Every speaker it finds is listed, each tagged to show whether it matches an existing host or will be created as a new one.
4. Import. Jellypod matches speakers to your podcast's regular hosts first, then the rest of your account, and creates a new host for anyone left, picking a voice in the script's language and writing a backstory from their lines.

Note: Newly created hosts don't impact your podcast's regular hosts and only apply for this specific episode. When this happens, a banner above the script editor will allow you to "Update Podcast" to sync your podcast's default cast to these hosts.

Your AI agent can now make podcasts for you.

We've shipped a Jellypod skill that teaches an agent how to use your Jellypod account end to end: design hosts, attach sources, draft the script, and render audio against the real Jellypod API.

Hand the agent a prompt like "spin up a weekly podcast that summarizes my company's product updates" and it picks the voices, builds the hosts, pulls in your sources, and produces the episode in your library.

The skill is open source. Install instructions for your agent of choice (Claude Code, Claude.ai, ChatGPT, and others) are in the repo: github.com/Jellypod-Inc/skills.

We just added 30 new voices to the Jellypod Voice Library.

They sound more natural, more expressive, and more flexible across different podcast styles. Some episodes need a calm narrator. Others need a sharper editorial voice, a warmer co-host, or a more energetic presenter who can keep the conversation moving.

To try them, create a new host or edit an existing one, choose Voice Library, and browse the first three pages. Give a few samples a listen, compare the delivery, and pick the voice that feels right for your next episode.

They are live now. Give them a try the next time you are casting a host.

Jellypod now has a dark mode.

Use the theme toggle in the top right corner of the studio to switch between light and dark. Your choice is remembered the next time you sign in.

For all those late-night editing sessions.

You write the prompt once. Jellypod takes it from there.

Automations let you put any podcast on a recurring schedule — daily or weekly — and have a fresh episode generated on its own. Each run kicks off an agent that researches the topic, outlines the script, generates the audio, and either drops it in your drafts for review or publishes straight to your feed.

It's the same pipeline that powers chat-based generation, just running for you on a clock.

What you configure

• Prompt — what each episode should cover. e.g. "Cover the most important AI product launches and research from the past day."
• Podcast — which show the episode posts to. Hosts and voices are inherited automatically.
• Schedule — daily or weekly, the days of the week, time of day, and timezone. A live preview shows the next three runs.
• Episode length — pick a duration preset.
• Web Search — on by default, so the agent can pull current information before writing.
• Auto-Publish — off by default. Runs save as drafts you can review, edit, and publish. Flip it on to ship every episode automatically.

What happens on every run

When the scheduled time hits, Jellypod spins up the episode in the background. You get an email when the draft is ready (or, with Auto-Publish on, when it's live). If a run fails, we email you with the reason and try again on the next tick. After three consecutive failures, we auto-pause the schedule so it doesn't burn credits silently — same if your account runs out of credits.

You can pause, resume, edit the prompt, or change the schedule any time from the Automations page in the sidebar.

Why we built it

Most podcasters who run a daily or weekly show are doing the same setup work over and over: open the app, paste the same prompt, generate, publish. Automations collapse that into a one-time setup. It's especially useful for news shows, recap formats, and any series where the topic is consistent but the content changes daily.

Find it on the Automations page in the sidebar.

Consistent publishing is also one of the most important factors in podcast monetization. See how to make money podcasting for a breakdown of which revenue models become viable as your show publishes reliably over time.

Automations is also the production step in a larger content marketing automation pipeline: wire it to a schedule, an inbox, or a trigger from the rest of your stack, and audio becomes a channel that runs itself instead of a project someone has to remember to start.

You can now add audio to any episode without leaving the transcript.

Drag a file from your desktop to insert in between transcript blocks. Or type /external audio in any empty speak block to open the Insert Audio dialog — record a new clip with your mic or upload a file. Either way, the clip becomes an inline block you can play right inside the transcript.

Use it for an intro you recorded yourself, a guest's voice memo, an interview snippet etc. You can even have a conversation with one of your hosts!

You can now programmatically create podcasts via our public API.

The Jellypod API enables clients to create AI hosts, upload sources, and generate full podcast episodes from a single prompt. Clients can browse all of our 100+ voices across 70+ languages, build hosts with custom backstories, attach URLs, PDFs, or YouTube videos as research context, and generate episodes asynchronously with a single request.

Using the Jellypod API uses credits equivalent to the UI (reference). You can also give your AI assistant the ability to generate podcasts by installing our Jellypod skill here.

Full documentation and an interactive reference are available at https://www.jellypod.com/docs/api.

You can now view and redownload all previous renders from your account settings.

The new Exports tab in your account Settings enables you to browse every render you’ve ever created. Each entry shows the title, type, date, and credit cost at a glance.

Need to grab a file again? Download the video, audio, or captions from any past render without spending additional credits. You’re only charged once, when the render is created. Re-downloads are always free.

Sometimes you want to go fast. Turbo Mode skips clarification and jumps straight to podcast episode generation.

Turbo Mode is now how every episode gets made. Describe your episode, hit send, and Jellypod makes reasonable assumptions based on your sources, builds the outline behind the scenes, and starts generating your podcast episode immediately. No clarification questions, no back-and-forth, and no required approvals.

Once generation finishes, review the script and audio and ask for changes, like "make the intro shorter" or "add a section about X", and the agent revises without starting over.