Which wallets can send ZEC to a TEX (tex1) address? Where things stand and what's left to fix

Hi all,

If you’ve tried to send ZEC to Binance lately, you may have run into this. Binance gives every ZEC user a deposit address that starts with tex1, and a lot of wallets reject it as “invalid address”. THORChain’s ZEC inbound address is often a TEX address too, so the same wallets can’t swap ZEC through THORChain either. The address is fine. The wallet just doesn’t support TEX addresses (ZIP 320) yet.

I’ve spent the last few weeks reading wallet code and filing issues, so here’s where things stand today (18 Nov). It comes from code and release notes, not from sending real funds out of every wallet, so please correct me if I’ve got one wrong.

Wallets that can send to a TEX address

Zodl, Zingo, Zkool, Edge, Unstoppable, Zallet, Gem, Vultisig, OneKey, KeepKey, Coin Wallet, Vizor, Zafu, Noir, Zapp, ZChat and Laurel.

One thing worth knowing: Zodl can pay a TEX address from a shielded balance (it does the two steps for you), but not from a Keystone or Ledger account yet.

Wallets that can’t yet (each has an open issue)

  • Cake Wallet: cake-tech/cake_wallet#3871
  • Brave Wallet: brave/brave-browser#61975
  • Ledger Live: LedgerHQ/ledger-live#25091
  • Trezor: trezor/trezor-firmware#4581 and trezor/trezor-suite#35232
  • Keystone: KeystoneHQ/keystone3-firmware#2313
  • ShapeShift: the fix is upstream, in christsim/multicoin-address-validator#20
  • SSP Wallet: already has a fix in review (RunOnFlux/ssp-wallet#770). That one came together in a day, so thanks to their team.

What’s been fixed so far

  • Zebra now shows a TEX address as tex1… instead of its t1… form (ZcashFoundation/zebra#11880), and on next it rejects a TEX address with the wrong kind of checksum (#12360, from #12254).
  • zecblock can now look up a TEX address (Kenbak/zecblock#284).
  • Zkool’s scanner now finds the temporary addresses it uses when paying a TEX address from shielded funds (hhanh00/zkool2#1639).
  • blockchain_utils, a Dart library, now rejects TEX addresses that carry the wrong amount of data. Mohsen fixed it within an hour of the report (mrtnetwork/blockchain_utils#19).

Two edge cases are now settled

ZIP 320 didn’t spell out two cases, and the ZIP 320 Owners have now decided both (zcash/zips#1468, with the ZIP change in zcash/zips#1469):

  • An address written in all capitals (TEX1…, which QR codes often use) must be accepted. Mixed case is still invalid.
  • Extra padding at the end of the encoded data must be rejected.

zcash_address and Zebra don’t handle either one yet (zcash/librustzcash#3306 and #3223, ZcashFoundation/zebra#12517). If your wallet decodes TEX addresses with its own code, it’s worth checking both.

If you build a wallet or an exchange

I put every bad address that caught a real bug this year into one test file, with the network each one should be checked on. It’s MIT and runs on plain Python: GitHub - cryppit/zcash-tex-vectors: ZIP 320 TEX address test vectors, a reference codec and an integrator checklist. · GitHub

The same invalid cases are proposed for zcash-test-vectors in zcash/zcash-test-vectors#139. The ones that tripped up real code were a Bech32 checksum where Bech32m is required, a 19- or 21-byte payload, an extra padding group, and a valid TEX address for the wrong network.

I’ve also asked several exchanges and closed-source wallets directly. None has told me it can send to TEX yet. If you work on a wallet or exchange that isn’t listed here, a quick reply saying whether it can pay a tex1 address would help a lot of people.

Disclosure: I maintain that test file and filed most of the issues above. This TEX work is part of our retro grant application (t/58901).

1 Like

Adding the exchange side, since that’s where most people get stuck. Moving ZEC from an exchange to a Binance deposit address only works if the sending exchange can pay a tex1 address.

  • Binance gives every ZEC user a tex1 deposit address (all of them since 1 Jan 2025) and only takes transparent ZEC there.
  • Coinbase can send to TEX addresses. Coinbase Assets announced it in October 2024.
  • Gemini is reported to support TEX withdrawals, but I couldn’t find Gemini’s own statement, so treat that as unconfirmed.
  • Kraken lists only t addresses in its supported address formats, and its support told me today that TEX isn’t supported yet and that they’ve passed the request to their team. So Kraken to Binance doesn’t work for ZEC right now.
  • Asked today, no answer yet: OKX, KuCoin, HTX, Poloniex, Bitget.
  • No public word either way: Gate, WEEX, Toobit, XT.COM, Bitfinex, WhiteBIT, BingX, Backpack, and Robinhood (which only says it allows withdrawals to “unshielded” addresses).

If you have a contact at any of these, pointing them to this thread would really help. The change on their side is small: accept tex1 in the ZEC withdrawal form and pay the P2PKH script of its 20-byte key hash from transparent funds. The test file in my first post has the valid and invalid cases they’d need.

As someone who coordinated the whole ecosystem implementation as part of my ecosystem developer grant I must say the use of TEX addresses should not be encouraged. They don’t solve a real problem. Their origin was not justified properly. no other exchange requires them. Community felt obliged to given a low price action and what a delisting from Binance could mean at that time.

If you build exchange you don’t need to support TEX addresses at all. Just use the protocol as it is right now as 99.99% of your competitors do.

That being said.

It is a very BAD idea to withdraw from a DEX to a CEX directly. Always send your ZEC to a wallet you fully control and then send to your exchange of choice. Otherwise in case of needing refunding you’ll be in trouble.

Thanks, that’s useful context, and fair.

To be clear about what I’m asking for: I’m not suggesting any exchange should start issuing TEX deposit addresses, and I agree there’s no need to. The ask is narrower. Because Binance does issue them, a wallet or exchange that sends ZEC should be able to pay one, so people don’t get stuck on an “invalid address” error. That’s the only thing the lists above are about.

I agree on DEX to CEX too. Sending a swap’s output straight to an exchange deposit address is how refunds go wrong. Sending to a wallet you control first, then on to the exchange, is the safer route, and it’s what we’ve been telling people on X as well.

With all due respect, I don’t agree. I can totally understand if product developers from different exchanges see tex addresses as Binance addresses and deliberately choose not to support withdrawing to them. It’s a product choice, eventually their customers will choose whether that was the right call or not. TEX addresses are not a consensus rule and it was on purpose that they were not. :slight_smile:

1 Like

Fair point, and you know that history far better than I do. It’s their product call.

One thing that would help users either way: when a wallet or exchange decides not to pay TEX, saying so instead of “invalid address”. Something like “This is a TEX address (ZIP 320). We don’t send to these; use a wallet that does.” Then the customer knows the address is fine and can make the choice you mention, instead of thinking they copied it wrong.

I’ll add that to the checklist in our test file. Do you think it’s worth a recommendation anywhere else, or is that also best left to each product?

I think that User Experience recommendations are something you could propose as part of ZIP-320.

1 Like

If ZIP 320 gets the UX section pacu suggested, what exact words should a wallet show when it can’t pay a tex1 address yet, so a new user doesn’t throw away a good deposit address?

Zcash wallets and software are different levels of opinionated and have been for years. We do have over 30 different wallets now so finding a suitable one that works in agreeance with your needs or principles or whatever else should be feasible.