Lockally

      Use your API key in the examples

      Replaces lk_live_… in the code samples on this site so you can copy and run them. It is stored in this browser only and is never sent anywhere, including to Ask AI.

      It will be visible on the page. Use a test key if you are sharing your screen or taking screenshots.

      Sender identity

      The from address is the one field the API will not take on trust. What counts as a valid sender depends on your account mode, and the two rules exist for different reasons.

      The rule, in one table

      Mode You may send from
      API Any address on a domain you have verified. The address need not exist.
      Business, Both A mailbox you hold, or an alias on a verified domain.

      Anything else is rejected with 403, before the message is queued and before anything counts against your allowance.

      Why API accounts do not need the address to exist

      An API account cannot create mailboxes at all: POST /v1/mailboxes is gated to business mode. If sending required a mailbox, every API account would have zero usable sending identities, which is not a rule so much as a locked door.

      Domain verification is the proof that actually matters. It is what DKIM signs, what SPF authorises and what the return path is built from. A mailbox row would add nothing to the authentication story; it would only add a billed seat and a storage quota for an address nobody will ever open.

      So on a verified domain, noreply@, receipts@, alerts@ and anything else work the moment verification completes, with nothing to create first.

      Why business accounts may use an alias

      Until October 2026 business and both accounts had to send from a real mailbox. Aliases were excluded on the reasoning that these accounts can create mailboxes, so the requirement cost them nothing.

      That reasoning was wrong, and it is worth saying why rather than quietly changing it. A mailbox is a billed seat. The rule charged a yearly seat fee for the privilege of sending as no-reply@ or security@, addresses that exist to send and never to receive. The honest answer we had to give customers who asked was “buy more mailboxes”, which is a poor thing to say to somebody deciding whether to pay at all.

      An alias is safe for exactly the same reason API mode is: an alias cannot be created unless its domain is already tenant-owned and verified, so the proof DKIM signs is in place before the alias exists.

      A catch-all does not widen this. Catch-all lives on the domain record and creates no alias rows, so it does not confer a sending identity.

      Verification does not inherit

      Verifying example.com does not authorise mail.example.com, or any other subdomain. Each is proved separately, because each is a separate DNS zone from the point of view of the records that matter.

      This is the single most common 403 on a first send: the domain looks verified in the dashboard, but the address is on a subdomain nobody verified.

      What the recipient actually sees

      Two things decide the name next to your address, and neither is the from field:

      • Business mailboxes show the mailbox’s own display name, which its owner controls.
      • Everything else falls back to the address itself.

      You cannot set a display name through from on POST /v1/send: the field takes a bare local@domain address and is rejected otherwise. Use reply_to_name to name the reply address, which is the part a recipient interacts with.

      Checking before you send

      There is no “can I send as this address” endpoint. The practical check is GET /v1/domains, which lists your domains and their verification state, and GET /v1/mailboxes or GET /v1/aliases on a business account.

      A send that fails this gate returns 403 with the title Sender not authorised, and the detail explains the rule for your mode.

      It carries no code, so there is nothing stable to branch on beyond the status. business_mode_required is a different thing: it comes from endpoints that are closed to API accounts entirely, not from the from field. If you need to tell “wrong sender” apart from other 403s programmatically, that is a documented gap rather than something you can do today.