UPDATE: even if it is an old question, here is the latest update.
It is now possible using the v1alpha2 version. The gRPC message AssistConfig
is defined as an union where you can choose between an audio out config or a text query.
If you are using the python library, see the AssistConfig
. Here is an example of config using a text query (adapted from the pushtotalk.py
sample, line 183):
config = embedded_assistant_pb2.AssistConfig(
# instead of audio_in_config
# note: no need to use 'OK google'
text_query = "who are you ?",
audio_out_config=embedded_assistant_pb2.AudioOutConfig(
encoding='LINEAR16',
sample_rate_hertz=self.conversation_stream.sample_rate,
volume_percentage=self.conversation_stream.volume_percentage,
),
dialog_state_in=dialog_state_in,
device_config=embedded_assistant_pb2.DeviceConfig(
device_id=self.device_id,
device_model_id=self.device_model_id,
)
)
If you are using golang, here is the link to the godoc.