Skip to content

openai_text_to_speech_driver

OpenAiTextToSpeechDriver

Bases: BaseTextToSpeechDriver

Source code in griptape/drivers/text_to_speech/openai_text_to_speech_driver.py 
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
@define
class OpenAiTextToSpeechDriver(BaseTextToSpeechDriver):
    model: str = field(default="tts-1", kw_only=True, metadata={"serializable": True})
    voice: Literal["alloy", "echo", "fable", "onyx", "nova", "shimmer"] = field(
        default="alloy",
        kw_only=True,
        metadata={"serializable": True},
    )
    format: Literal["mp3", "opus", "aac", "flac"] = field(default="mp3", kw_only=True, metadata={"serializable": True})
    api_type: Optional[str] = field(default=openai.api_type, kw_only=True)
    api_version: Optional[str] = field(default=openai.api_version, kw_only=True, metadata={"serializable": True})
    base_url: Optional[str] = field(default=None, kw_only=True, metadata={"serializable": True})
    api_key: Optional[str] = field(default=None, kw_only=True)
    organization: Optional[str] = field(default=openai.organization, kw_only=True, metadata={"serializable": True})
    _client: openai.OpenAI = field(default=None, kw_only=True, alias="client", metadata={"serializable": False})

    @lazy_property()
    def client(self) -> openai.OpenAI:
        return openai.OpenAI(
            api_key=self.api_key,
            base_url=self.base_url,
            organization=self.organization,
        )

    def try_text_to_audio(self, prompts: list[str]) -> AudioArtifact:
        response = self.client.audio.speech.create(
            input=". ".join(prompts),
            voice=self.voice,
            model=self.model,
            response_format=self.format,
        )

        return AudioArtifact(value=response.content, format=self.format)

api_key: Optional[str] = field(default=None, kw_only=True) class-attribute instance-attribute

api_type: Optional[str] = field(default=openai.api_type, kw_only=True) class-attribute instance-attribute

api_version: Optional[str] = field(default=openai.api_version, kw_only=True, metadata={'serializable': True}) class-attribute instance-attribute

base_url: Optional[str] = field(default=None, kw_only=True, metadata={'serializable': True}) class-attribute instance-attribute

format: Literal['mp3', 'opus', 'aac', 'flac'] = field(default='mp3', kw_only=True, metadata={'serializable': True}) class-attribute instance-attribute

model: str = field(default='tts-1', kw_only=True, metadata={'serializable': True}) class-attribute instance-attribute

organization: Optional[str] = field(default=openai.organization, kw_only=True, metadata={'serializable': True}) class-attribute instance-attribute

voice: Literal['alloy', 'echo', 'fable', 'onyx', 'nova', 'shimmer'] = field(default='alloy', kw_only=True, metadata={'serializable': True}) class-attribute instance-attribute

client()

Source code in griptape/drivers/text_to_speech/openai_text_to_speech_driver.py 
29
30
31
32
33
34
35
@lazy_property()
def client(self) -> openai.OpenAI:
    return openai.OpenAI(
        api_key=self.api_key,
        base_url=self.base_url,
        organization=self.organization,
    )

try_text_to_audio(prompts)

Source code in griptape/drivers/text_to_speech/openai_text_to_speech_driver.py 
37
38
39
40
41
42
43
44
45
def try_text_to_audio(self, prompts: list[str]) -> AudioArtifact:
    response = self.client.audio.speech.create(
        input=". ".join(prompts),
        voice=self.voice,
        model=self.model,
        response_format=self.format,
    )

    return AudioArtifact(value=response.content, format=self.format)