- Nom d'utilisateur
- bestminerineurope
- Adresse du portefeuille
- bc1qecyz...emyvlp03
- Date d'inscription
- Total des plages
- 0
- Total des messages
- 8
New account, first post here. I want to share with you a dataset I spent several months building but I feel it's not a right idea to give it for free due to the effort spent to collect it. Free sample below — judge it before donating anything.
What it is
There is a YouTube channel called "Bitcoin Puzzle 71 Prefixes" that displays private keys in the #71 range whose addresses share a leading prefix with the puzzle target. The keys are flashed far too fast for a human to read and the frames are visually distorted, apparently on purpose.
I recovered them anyway:
- Downloaded every video on the channel
- Extracted all frames
- GPU-accelerated OCR with preprocessing to handle the distortion
- Deduplicated and cleaned
Result: ~1,000,000 rows of <71-bit hex key>,<address>, all with the 1PWo3JeB prefix, all inside the #71 range.
What I am NOT claiming — read this before buying
I do not believe this data helps solve puzzle #71, and I'm not going to imply otherwise to make a sale.
Addresses are Base58Check(RIPEMD160(SHA256(pubkey))). Hashing destroys structure, so two keys whose addresses share a prefix have no mathematical relationship to each other. The matching keys are uniformly scattered across the range. A million samples from that set carries no information about which candidate the puzzle creator chose.
And since #135 came up recently. #71 has no exposed pubkey, so it remains pure brute force over 2^70 with a hash per candidate. Prefix data does not shortcut that.
So what am I actually charging for?
Labour and compute, not a lottery ticket. Several months of GPU time, frame extraction across the channel's full back catalogue, and an OCR pipeline built specifically to defeat the distortion the uploader applied. The dataset does not exist anywhere else because nobody else has done this work. That is what the price reflects. If that isn't worth it to you, don't donate in the wallet below — I'd rather have no donations than a community member who expected something else.
What you might actually want it for
- Testing whether the channel's key generator is uniform or biased. That is a fact about the generator, not the puzzle, and it's the only genuinely open question in here as far as I can tell.
- A ~1M-row exclusion list of in-range keys confirmed not to be the answer.
- Sample data for statistical work on the range. (if you have a data science background)
Free sample
100 rows, no conditions: [https://ufile.io/skrkixo2] (Files on ufile.io will expire on 7 September 2026)
Verify it yourself. Don't take my number on faith.
Terms
The idea is simple: as a community, you can contribute any amount you like—from $10 on up—to our designated wallet.
Wallet address => bc1qg2ay6jvcvqd4f8r35acp7738932s6ulssuzq0u
Once the total balance in the wallet reaches our community goal of $500 USD, I will immediately release and share the Full Database with everyone.
This way, the cost is shared, and we unlock the resource together as soon as the target is met!
Happy to answer anything about the extraction pipeline or the verification method.
PS: These prefixes are often called "BS prefixes" for a reason, so think twice before sending anything. Let's see what happens :)