139 139
140When using the Speech to Speech API over SIP, phone keypresses (DTMF tones) are automatically buffered and flushed to the model as text input. The client receives `input_audio_buffer.dtmf_event_received` events as an audit trail of each keypress.140When using the Speech to Speech API over SIP, phone keypresses (DTMF tones) are automatically buffered and flushed to the model as text input. The client receives `input_audio_buffer.dtmf_event_received` events as an audit trail of each keypress.
141 141
142To let the agent press keys itself, for example to navigate a phone menu, add `{"type": "function", "name": "send_dtmf"}` without `parameters` to `tools` in `session.update`; the API supplies the schema and sends the tones, so do not reply to `send_dtmf` function calls.
143
142### Flush triggers144### Flush triggers
143 145
144Buffered digits are submitted to the model when any of the following occurs:146Buffered digits are submitted to the model when any of the following occurs: