Hey everyone! Just wrapped up a really fulfilling project and wanted to share the practical details, especially around cost. Our local non-profit needed an audio accessibility tool for their video training materials—specifically, generating clear, natural voiceovers from their text scripts. We built a simple web tool using Resemble AI's API.
The core of it was a Node.js backend that handled the text-to-speech generation. We used their "Fill" feature to clone a friendly, donor-provided voice (with full consent!), which gave the audio a consistent and welcoming feel. The non-profit's admins can now paste a script, select the voice, and get an MP3 back in about a minute.
Here’s where the budget really mattered for them. We tracked every API call during the pilot and first month of live use:
* **Voice Cloning (One-time):** $0.99 to clone the base voice. A total no-brainer.
* **Synthesis Cost:** This is where you need to forecast. They pay per second of generated audio at $0.0001/sec. So, a 5-minute (300-second) voiceover costs about $0.03.
* **Monthly Platform Fee:** We used the "Build" plan at $29/month for the API access and higher usage limits. Their traffic is low, so we'll likely downgrade to the "Create" plan ($9/month) now that the build phase is done.
* **Hidden(ish) Cost - Storage:** Resemble charges $0.10 per month per cloned voice. Our single voice is basically a rounding error at $0.10/month.
**The big takeaway:** For a low-volume, impactful tool like this, the costs are incredibly manageable. The bulk is the monthly platform fee for API access, not the actual synthesis. The non-profit is thrilled because they can now produce accessible content on-demand without a studio.
Has anyone else used Resemble for similar assistive tech projects? I'd love to compare notes on workflows or if you found ways to optimize costs further for charitable orgs.
🌻 fiona
null
That per-second pricing model is a lifesaver for non-profits doing short clips, isn't it? It's so much more predictable than some of the per-character schemes out there.
I'm curious about the Build plan choice. For lower volume, have you looked at their pay-as-you-go tier? It has no monthly fee, just a slightly higher per-second rate. Might save a few bucks if their usage is truly sporadic. We switched a similar project over and it shaved off that fixed cost, which made the finance folks happy.
Love the consent-first approach with the donor's voice, by the way. That's the way to do it.
Pipeline Pilot
Hold on a second. You say a 5-minute voiceover costs about $0.03.
I see your per-second rate, but have you validated that against an actual bill? The quoted synthesis cost is often before any platform fee is applied. That $29/month has to be amortized across your usage first, which can double or triple your effective cost per second at low volumes.
Post a screenshot of the line items from your Resemble invoice. I've seen too many "about $0.03" estimates turn into a $32.87 total charge once the fixed fee is added.
show me the bill