Audio for a tip
The clip for a queued tip, made on demand. A standard-voice tip’s audio is only made when something comes to play it, so a consumer that plays tips itself asks here before or after claiming; an AI-voice tip’s clip was made at payment and is handed straight back. The words come from the row’s current name, amount and message. The clip you are handed is written to the queue row as tts_url and fires tip.tts_ready (again if it replaces an earlier clip in another voice, such as a dashboard preview).
Say whose voice to read it in: overlay for that overlay’s alert box voice, or voice for a standard voice. Passing neither is deprecated: it reads in Brian for now and will become 400 voice_required. Other voice parameters given directly override the overlay’s. A tip that has already played can be asked for too: you get the clip that played, whatever voice you name (one that played with no audio is made now). No rate limit; up to 8 distinct voice variants per tip. The same parameters are accepted in a JSON body on POST. Requires tts:read.
/tts/{orderId}/audioAuthorizationBearer token (tp_live_... or tpat_...) · headerrequiredA TipPage API key (tp_live_..., created in Dashboard → Settings → Developer) or an OAuth 2.0 access token (tpat_..., see the OAuth guide). Both carry the same scopes and are accepted on every endpoint.
orderIdstringrequiredThe tip's order id (e.g. tip_1755100000000_ab12cd) - the only external row reference.
overlaystringAn overlay id (o_...). Read the tip the way that overlay's alert box is set to.
voicestringA standard voice id from /tts/voices.
ratenumberpitchnumberreadNamesbooleanreadAmountsbooleanreadMessagesbooleanmaxLengthintegerCut the message at this many characters.
One of: the clip (url, with text for a standard tip or ai for an AI-voice tip), pending (an AI voice is still being made, ask again after retryMs), or silent (nothing of this tip is read with these settings).
urlstring<uri>textstringThe words in the clip.
cachedbooleanproviderstring | nullaiobjectPresent for an AI-voice tip.
Show propertiesHide properties
voicedboolean | nullFalse when the AI generation failed and the clip is the standard voice.
turnsobject[] | nullpendingbooleantrueretryMsintegermessagestringsilentbooleantrueA voice parameter is out of range or names no standard voice. Later also voice_required (neither overlay nor voice was given).
errorstringrequiredHuman-readable message.
codestringMachine-readable code (e.g. missing_scope, not_in_queue).
Missing or invalid credential (missing_api_key / invalid_api_key, or invalid_token for an expired/revoked OAuth access token).
errorstringrequiredHuman-readable message.
codestringMachine-readable code (e.g. missing_scope, not_in_queue).
The key lacks the required scope (missing_scope).
errorstringrequiredHuman-readable message.
codestringMachine-readable code (e.g. missing_scope, not_in_queue).
required_scopestringtip_not_found (no tip of yours has that id) or overlay_not_found (no overlay of yours with an alert box has that id).
errorstringrequiredHuman-readable message.
codestringMachine-readable code (e.g. missing_scope, not_in_queue).
too_many_variants: the tip has been asked for in too many different voices.
errorstringrequiredHuman-readable message.
codestringMachine-readable code (e.g. missing_scope, not_in_queue).
Every synthesiser failed. text carries the words, for a voice of your own.
fallbackbooleantruetextstringerrorstring