Introducing Horizon: Our Next-Gen Voice Model
Today we're releasing Horizon, our newest and most advanced AI podcast audio model. Horizon excels at delivering higher fidelity audio, richer expressions, and more natural dialogue and conversation.
It's designed to make your host sound more lifelike and more engaging, with clearer diction, better pacing, and subtle emotional cues that bring conversations to life.
All Jellypod users get access to Horizon at no additional cost.
How do I turn it on?
By default, all new users have Horizon chosen as their default audio model. For existing users, you can turn on Horizon under your account settings.
If at any point you prefer the previous model, you can revert to our Classic model in your Account Settings. This may be applicable if you're used to a voice clone sounding a specific way.
Note: Due to it's higher quality and realism, audio generations may take longer than our Classic.
Tips for Perfect Audio
Horizon is more sensitive to context and emotion. If a host sounds too animated or too flat, try these quick adjustments:
- Refine the backstory: Add guidance about tone, energy level, or demeanor. Example: "Keeps a calm, measured delivery."
- Adjust block speed: In the editor, tweak the speed for any block that feels too fast or too slow.
- Balance the cast: If you bring in a high-energy host, pair them with a steadier co-host for contrast.
FAQs
- Will Horizon change my existing episodes?
- No. Previously published episodes are not altered.
- What if I hear over-the-top emotion?
- Tone it down by softening adjectives in the backstory and removing exclamation marks in scripts. You can also lower speed a touch for a calmer read.
- My voice clone no longer sounds like me! Help!
- If the new voice is unacceptable, you can always revert to our Classic model in your settings.
What's Coming Next
We will roll out several new Horizon-specific features in the coming weeks, giving you even more fine grained control over how your hosts sound. Stay tuned!
