Crypto Bounty Intelligence

Unsolved on-chain cryptographic bounties — verified reference dossier · Coastal Consulting · Oct 2026

The 1000 BTC Challenge transaction verified on-chain

The 2015 "Bitcoin Puzzle" transaction 08389f34... remains the largest active on-chain cryptographic bounty pool: 916 BTC unspent across its unsolved outputs. Puzzles #1–#70 are solved; the active sequential front is puzzle #71 (7.1 BTC, ≈$607,574), keyspace [2^70, 2^71 − 1] ≈ 1.18×10^21 keys. No public key is exposed for #71, so solvers must run full EC point multiplication + SHA-256 + RIPEMD-160 against the target address. Coordinated pools (btcpuzzle.info and others) actively partition the range.

Kangaroo-tractable targets (public keys exposed) O(√N)

Puzzles #135, #140, #145, #150, #155, #160 (13.5–16.0 BTC each) had uncompressed secp256k1 public keys exposed on-chain via historical spends/reuse. Knowing P = k·G unlocks Pollard's Kangaroo / BSGS, cutting complexity from O(N) to O(√N). Tooling: theCollider (CUDA/Metal), Keyhunt (CPU/BSGS), DocKangaroo (containerized distributed nodes), BitCrack (CUDA/OpenCL).

Standalone escrows & steganographic bounties

Target	Bounty	Vector
Bitcointalk 1 BTC	1.000 BTC	Unspent since 2019; forum clue chain
GSMG.io Final Gate	1.256 BTC	5-phase decode; halted at sealed AES-256-CBC blobs
Bitaps SSSS Challenge	1.000 BTC	Shamir's Secret Sharing quorum → BIP39
Bobby Lee Ballet Cards	2.000 BTC	Physical cold storage; photo entropy missing
Aoi Nakamoto Quizchain	0.777 BTC	Literary mining → MD5 → key
Peter Todd bounties	~0.594 BTC	Truncated-hash collision challenges
BLM Steganography Collage	0.200 BTC	Micro-runes, dial coords, pixel LSB
Trithemius: Wealth in Poetry	0.0312 BTC	Classical cipher in poetic meter
Corey Phillips Kitten	0.0100 BTC	Known 12-word seed; passphrase brute force
School of Bitcoin	0.0100 BTC	Image payload → key bytes
RushWallet Contest #30	0.0100 BTC	Brainwallet passphrase
Zden Level HALV	0.0031 BTC	Visual geometric series

Tracked by open repositories: agntn/puzzles, floflo777/open-crypto-puzzles, roadhero/Bitcoin-Puzzle-Info.

Part II — The obscure front: 28 funded puzzles the big trackers miss verified Oct 2026

Beyond the famous pools sits a long tail of funded, unsolved, rarely-cited escrows tracked by floflo777/open-crypto-puzzles (open research repo, escrows re-checked on-chain Aug–Sep 2026) and agntn/puzzles. Totals across the tracker's 28 funded unsolved puzzles (one, LogicBeach, fell on Oct 5, 2026 — see below): 5.94 BTC, 13.21 ETH, 1,900 AR, 3.03 LTC, 306 USDT ≈ $403,000.

Live spot-check (Oct 6, 2026): the Keir Finlow-Bates 12-lot book hunt shows 9 lots swept by solvers, 3 still funded (200,000 sats each, verified on mempool.space). Proof the method pays out — the small ones do fall.

Puzzle	Prize	What makes it crackable
Guntis Vitolins: 10 ETH Challenge	8.61 ETH	12-word seed scattered across the author's own video-series clues; bounded compute
TeikhosBounty (Johan Nygren)	2.00 ETH	Proof-of-public-key puzzles held in a smart-contract timelock
Keir Finlow-Bates: book treasure hunt	600k sats	Key hidden in a published book's text; 3 lots open, write-ups exist for the 9 solved
LogicBeach: Powerful Moss	0.55 ETH	SOLVED — see below
Bountiful: Fe compiler bug bounty	1.00 ETH	Not a cipher — a real compiler-bug hunt, escrowed in a timelock contract
Arweave Puzzles #3 / #10 / #11 / #12	~2,800 AR	Word-selection, geometry and pixel ciphers; #11 is a raw 64-hex key in one grayscale sketch
AH White: Walking Banks	800k sats	BIP39 seed encoded in a book's text cipher
Smith, Lyle & Moore: Glimmer	0.032 BTC	Tree of password-locked web pages; each opened page hands you a seed fragment
Path to Greatness	3.03 LTC	Author's own cipher drawn in a clue image, with a microsecond-fast correctness test
FTPK Season 2: Never-Ending	306 USDT	BIP39 word-selection, USDT escrow
Pindar Van Arman: cryptoArtGAN Act 1	1 NFT	BIP39 seed via AI-collaborative art
Genesis Block Wallet Puzzle	169k sats	Multisig + raw private key puzzle
Crypto Puzzles 2018: #2	0.05 ETH	Image stego + video series, raw key
Exitonly Challenge 14	30k sats	BIP39 word-selection; prize below the effort floor — listed for completeness

The repo's maintainer reports working ~40 of these and cashing 0.5 ETH + ~0.01 BTC solo — that is the realistic shape of this niche: insight pays, compute doesn't.

Recently cracked — LogicBeach: Powerful Moss claimed Oct 5, 2026

On-chain forensics (run Oct 6): the winner wallet 0x6357...48f3 itself called withdraw() on the prize contract in Base block 52215961 at 2026-10-05 17:27:49 UTC, taking the full 0.553 ETH pot (wallet balance: 0.0064 → 0.5594 ETH). The funds were moved on within hours; the contract now holds 0. Some solver out-derived the published research grid (3 confirmed hours + 9 ambiguous × ~3 candidates, 3.3M combos already tested) — most likely by resolving the numeral-row ambiguity the way the open leads suspected. No public write-up yet. Claim tx on basescan.

Operating lesson, logged: always check escrow liveness before spending a single cycle on a puzzle. This check took 3 RPC calls and saved a pointless brute-force pass — the pot was swept ~26 hours before we started. The long tail is real, but it is also a competitive arena: the small escrows fall to whoever reads them first.

Scam perimeter

The long tail is full of imitators. Rules that hold:

Verify the escrow yourself on a public explorer before touching any puzzle. If the coins aren't sitting on-chain in the puzzle address, there is no puzzle.
Never pay for "hints" or "solutions" — every legitimate puzzle is free to attempt, and the prize is the payment.
Clone "1000 BTC puzzle" addresses and lookalike graphics circulate constantly; only the original 08389f34... transaction's outputs count.

Honest feasibility math read this first

None of these are a business plan. The numbers:

#71: 1.18×10^21 keys. A strong GPU does ~10^9–10^10 keys/s. Even pooling 1,000 GPUs (10^13 keys/s): ~3.7 years expected. The existing pools already race this — you'd be buying lottery odds against coordinated farms.
#135 kangaroo: ~2^67 ≈ 1.5×10^20 group ops. At 10^9 ops/s per GPU: ~4,700 machine-years. Infeasible at hobby scale.
The small escrows (0.2–3 BTC) are occasionally cracked by solo sleuths — with deep niche skills (steganalysis, classical cryptography, literary forensics), not brute compute. That's the only "edge" a solo operator realistically holds.

This page is published as a reference and intelligence dossier, not as an active mining operation. No cracking workloads run on this site or its owner's infrastructure.