The FastScribe
Audio to timestamped text. No signup, no cloud AI, no trace.
I run a transcription SaaS. The obvious build is a web form in front of a cloud speech API, and I started there on paper. The arithmetic killed it.
Per-minute API pricing means a free tier is a direct subsidy. Every free file is money out. So every competitor's free tier is either tiny, or gated behind an account so they can throttle you, or both. That is not greed, it is the cost structure showing through the pricing page.
I already had a server. I put a used 3090 Ti in it. Now the marginal cost of a transcription is electricity, which changes what I am able to offer: the first file needs no account at all, and 5 a day free with one. I am not subsidizing anything, I am using capacity I already pay for.
The measured part, since numbers are the only interesting bit of a story like this: moving transcription from CPU to that GPU came out at roughly 55x real-time on the model I use, and about 105x on a smaller one. A one-hour recording finishes in about a minute. On CPU it was slower than listening to it.
What this actually costs me, honestly:
- One machine. No redundancy. If it dies, the queue stops and I get an alert.
- I own the ops. Fan curves, ECC errors, thermal alerts. I have a cron job that texts me if the GPU gets hot. A cloud API user has none of this.
- It does not scale past this box without me buying another one, and I would have to want that.
What I would tell someone considering the same thing: this only works because the workload is batch, not interactive, and because I already had the machine and the ops habit. If you would be buying a server to avoid an API bill you have not yet received, do not. The API is the right call until the bill hurts.
The privacy consequence was accidental and turned out to be the best thing about it. Because there is no third-party API in the path, I can say the audio never reaches OpenAI, Google or AWS, and I delete it the moment the transcript exists.
Competitors whose business model needs an account cannot copy the no-signup claim, and competitors who forward audio to a cloud API cannot copy the privacy one. I did not plan that. I backed into it by being cheap.
The site is thefastscribe.com if you want to break it. Happy to go into the hardware or the queue design.
7 Comments
The interesting part isn't the GPU decision — it's the business model shift it enables.
What would convince you that owning the infrastructure creates a lasting advantage, rather than just lowering costs until competitors can match the same economics?
Electricity costs are always lower than API providers.
Interesting. I was actually asking from a slightly different angle, but I can see why cost is the first thing that comes to mind.
Fair point, let's step away from unit economics. For us, the shift is about data sovereignty as a product feature.
API renters are legally bound by a chain of third-party cloud agreements, hypervisors, and data processing boundaries they don't control which force them to ask customers for trust blindly. By owning the bare metal, we can guarantee zero data persistence and absolute isolation at the silicon level. The lasting advantage isn't cheaper transcription; it's a verifiable privacy guarantee that unlocks highly regulated enterprise markets that API-dependent competitors are structurally barred from entering.
That clarification helps. The distinction between lowering costs and creating a different trust proposition is interesting.
I'd like to continue the conversation outside the thread. What's the best email to reach you on?
“‘No signup, no cloud’ is a strong trust play. Have you seen that drive repeat usage, or is it more of a one-off tool for most?”
I see pretty consistent usage so far, but we will see!