delete

[Kulahin]

delete

[TryNinja]

You'll have to figure out the exact same order of each word in the seed. I'm not aware of any script that brute-forces each possible seed with the 18 words, but there should be one (or maybe someone can write one for you).

The xpub and address will only be useful to verify if the wallet you restored with each seed is the right one or not.

[pooya87]

if you don't know the sequence in which those coins should be in, then there is no way that you can find your keys even with brute forcing because there is just too many variations of it possible that makes brute forcing impossible.

even if you know the first word there still is 17 more to go and that is 355,687,428,096,000 different combinations that you need to check and we are talking about a long process of deriving the master public key of each of these which requires ECC calculations and it is the bottleneck and comparing it with what you've got.

[bob123]

Unfortunately, pooya87 is right with the fact that its not pactically doable.

3.5 x 10¹⁴ combination are way too much to be bruteforced. It is fine to permutate a few words in a set of words. But iterating through all combination with only knowig 1 position is too much to accomplish.

Mind telling us HOW you could get your words into a mixed order ? I heared from cases where people lost some words, but why are they in a mixed order ?
Did you mix them up with a second sheet containing the order (which is missin now) ? Just curious.

[...] and we are talking about a long process of deriving the master public key of each of these which requires ECC calculations and it is the bottleneck and comparing it with what you've got.

The 'good' thing would be that you don't need to do that calcuations for every seed.
Thanks to the checksum, most of the iterated seeds (90%+ ?) will be invalid. Only a relatively small percentage of these seeds has to be derived.

This optimizes the process by a lot.. but unfortunately it is still not enough to help OP to find the correct seed.

[HCP]

18 words? The "common" number of words in a seed mnemonic is usually 12 or 24. Does Coinomi only use 18 word seed mnemonics? or are you simply missing 6 words?

[angelitto74]

18 words? The "common" number of words in a seed mnemonic is usually 12 or 24. Does Coinomi only use 18 word seed mnemonics? or are you simply missing 6 words?

Seeds can be any multiple of 3. Even just 3 will work, we often use this for our tests, for ease and speed. (Not recommended of course for real use, as 3 words can be bruteforced easily).

Coinomi simply implements BIP32/39/44 technology, and the default seed given on a new installation was an option of 18 or 24 words. Latest releases removed the 18 words option, and give out 24 word seeds.

Coinomi uses the open source upstream project https://iancoleman.io/bip39/ (which was forked to https://www.coinomi.com/recovery-phrase-tool.html but now we uploaded all our coins and functionality back upstream, as the original project also includes SegWit derivation paths.)

You can use those tools offline. You can see that one can select any multiple of 3 words to generate a new random seed.

The OP saved the "confirmation" screen, which comes after the screen that presents the new seed, and shows *scrambled* words, so the user can confirm his sequence.

Saving a screenshot is wrong anyway, and worse, saving this screenshot on the phone, unecrypted, voids the whole point. User should write down the words in a physical paper and keep it safe, away from online devices.

Unfortunately as mentioned above it is computationally unfeasible to brute force a 18+ word seed.