I built a self-hosted Telegram subscription bot that accepts ETH — no Stripe, no PayPal
Been working on this for a few months. Tired of seeing creators get banned from Stripe or have PayPal freeze their funds.
AnonStream is a complete Telegram bot where subscribers pay in ETH to access premium video content. Videos stored on IPFS, so no central server can take them down.
What's included: weekly/monthly plans, 2h free trial for new users, 4-tier referral system, admin dashboard, broadcast messages, analytics.
Runs for ~$7/month on Render. You keep 99.5% of every payment.
Selling the source code on Gumroad ($59) and a done-for-you setup ($179). Happy to answer any questions about how it works.
Great build! I built something similar — Telegram subscription bot with Paystack integration for African markets. Different payment focus but same concept. Would love to compare notes!
Clever architecture, and the OnlyFans comparison is the right framing, but it is also the trap. Dropping Stripe, PayPal, or OnlyFans does not drop the compliance.
It moves onto whoever runs the bot, because those platforms were quietly doing it for creators.
The big one for this use case: if any of the content is adult (and the OnlyFans framing suggests many buyers will use it that way), US law brings 18 USC 2257 performer age-verification records, and a growing list of states now legally require age verification for adult sites (the Supreme Court upheld that in 2025). "Anonymous, on IPFS, nothing can be removed" is the exact opposite of what those rules want, and immutable storage of user-uploaded content is its own liability if something illegal ends up in it.
Paying through NowPayments helps, but the operator still owns the sanctions and tax exposure a card processor used to absorb. None of this kills it as a dev product.
The operators who survive are the ones who know the platform was their compliance layer, and documenting that for your buyers could even be a selling point.
Interesting approach. I'm curious, how did you validate demand for this before spending months building it? Did you already have creators asking for a solution like this?
No traditional validation, a client hired me to fix a similar bot he already had. While working on it together, I kept improving things beyond the original scope. Eventually I asked if I could build my own version from scratch. He said yes.
Having a real user with real pain points was enough signal for me.
Getting payments off traditional processors makes sense when creators are worried about account freezes.
One thing that gets interesting with ETH subscriptions is how retention changes when users have to actively manage wallet balances instead of card renewals.
Have you seen any patterns yet in trial users converting compared to a typical subscription setup?
Good point on the retention risk, it's real, and I thought about it early on. That's why I built in an automatic reminder 5 days before a subscription expires, so users get a heads-up to top up their wallet before it lapses.
As for data: it's early, but out of 9 real users, 5 converted to a paid plan ($10/week or $24/month). So roughly 55% conversion from trial, better than I expected honestly.
On the ETH anonymity point, you're right that ETH isn't fully anonymous on-chain. The system is designed to be modular though, so adding Monero support as an option is on the roadmap for users who need stronger privacy.
Still early days, but the signals are encouraging.