Descript’s New Pricing (September 2025) — Media Minutes & AI Credits Explained
Descript’s New Pricing (September 2025) — Media Minutes & AI Credits Explained
Descript has announced major changes to its pricing, rolling out on September 23, 2025. If you’ve been relying on Descript for your podcasting or content workflow, you’ve probably already seen talk about “media minutes” and “AI credits.”
The announcement has left a lot of creators scratching their heads. Let’s break down what’s actually changing, what it means in practice, and what podcasters should be thinking about.
What Changed?
Previously, Descript plans were built around transcription hours and usage caps tied to specific features. Now, those have been replaced by two shared resource pools:
- Media Minutes → debited when you upload or record audio, video, images, or screen captures.
- Example: Upload a 60-minute interview = 60 media minutes.
- Even still images count as 1 second each.
- Minutes don’t roll over month to month (though top-ups will be available).
- AI Credits → consumed when you use AI-powered features.
- Examples: Studio Sound (10 credits), filler word removal (10 credits), Eye Contact (10 credits), text-to-speech (~5 credits per minute), dubbing (15 credits per minute).
- Costs vary depending on the tool and even the model you choose.
- Credits don’t roll over month to month either.
In short: every upload or AI action now draws from a finite pool.
Plan Examples
Here’s a quick look at what the new plans include each month:
- Free → 60 media minutes, 100 one-time AI credits.
- Hobbyist → 600 media minutes, 400 AI credits.
- Creator → 1,800 media minutes, 800 AI credits.
- Business → 2,400 media minutes, 1,500 AI credits.
If you’re on a Legacy or Sunset plan, you can stay put for now, but Descript has said those plans will eventually migrate into the new system. You’ll get at least 30 days’ notice before that happens.
Why the Change?
To be fair, the shift makes sense from Descript’s perspective. AI features like Studio Sound, dubbing, avatars, and generative video require a lot of processing power, and usage-based pricing helps them cover those costs.
It also gives them flexibility to keep rolling out new AI features without locking themselves into fixed “unlimited” buckets. In theory, it’s future-proof.
Who Benefits, Who Doesn’t?
- Agencies and larger teams: The new system may not sting as much. They can spread costs across multiple seats, and they’re often more likely to experiment with AI features.
- Solo creators and small teams: This is where frustration is growing. If your workflow is mostly the basics — transcribing, trimming silences, removing filler words — you could end up paying more without gaining much extra value. Suddenly, the things that used to be “set it and forget it” are now capped.
Community Concerns
From the reaction so far, three themes keep coming up:
- Paying for features you may never touch.
Many podcasters just want clean transcripts and basic editing — not AI avatars or video regeneration. - Core reliability matters more.
Users report that transcription and sync still glitch sometimes. Having to also watch usage buckets adds another layer of stress. - Uncertainty about migrations.
Legacy users know change is coming, but aren’t sure exactly when or how much it will cost.
What Podcasters Should Do Now
- Audit your workflow. Look back at how many minutes you upload per month. Compare it to the new tiers.
- Decide which AI features you actually use. If it’s only a couple, credits might feel like overkill.
- Watch for migration notices. Legacy plans won’t last forever, so it’s worth planning ahead.
Alternatives to Explore
There aren’t a ton of strong replacements right now, but here are a few worth considering:
- Trebble → focused on simplicity for solo creators. It keeps to the essentials — transcription, silence/filler word removal, cleanup, plus an AI pass that flags rambling and repeats. Flat pricing, no credits or caps. (Full disclosure: Trebble is our product — we built it to stay simple, reliable, and affordable for independent podcasters and small teams.)
- Riverside → recording-first, with editing features that are catching up quickly.
- Audacity → free and reliable, but very manual.
- Adobe Podcast → solid for audio enhancement, though limited in scope.
Final Thoughts
Descript’s new pricing isn’t inherently “bad” — it makes sense for where the company is headed. But for many indie podcasters, it feels like paying more for things you don’t use while worrying about the basics.
Even if Descript eventually tweaks its approach (say, unlimited basics with credits for AI), the trust gap may remain. Once creators feel burned, they start looking for alternatives that better match their needs.
That’s the opportunity: to rethink your workflow and choose the tools that fit your priorities — whether that’s feature depth, affordability, or reliability.