@@ -103,66 +103,6 @@ def create(
103103 timeout : float | httpx .Timeout | None | NotGiven = not_given ,
104104 ) -> TranscriptionVerbose : ...
105105
106- model 's confidence in the transcription . `logprobs` only works with
107- response_format set to `json` and only with the models `gpt-4o-transcribe` and
108- `gpt-4o-mini-transcribe` . This field is not supported when using
109- `gpt-4o-transcribe-diarize` .
110-
111- known_speaker_names : Optional list of speaker names that correspond to the audio samples provided in
112- `known_speaker_references[]` . Each entry should be a short identifier (for
113- example `customer` or `agent` ). Up to 4 speakers are supported .
114-
115- known_speaker_references : Optional list of audio samples (as
116- [data URLs ](https :// developer .mozilla .org / en - US / docs / Web / HTTP / Basics_of_HTTP / Data_URLs ))
117- that contain known speaker references matching `known_speaker_names[]` . Each
118- sample must be between 2 and 10 seconds , and can use any of the same input audio
119- formats supported by `file` .
120-
121- language : The language of the input audio . Supplying the input language in
122- [ISO - 639 - 1 ](https :// en .wikipedia .org / wiki / List_of_ISO_639 - 1_ codes ) (e .g . `en` )
123- format will improve accuracy and latency .
124-
125- prompt : An optional text to guide the model 's style or continue a previous audio
126- segment . The
127- [prompt ](https :// platform .openai .com / docs / guides / speech - to - text #prompting)
128- should match the audio language . This field is not supported when using
129- `gpt-4o-transcribe-diarize` .
130-
131- response_format : The format of the output , in one of these options : `json` , `text` , `srt` ,
132- `verbose_json` , `vtt` , or `diarized_json` . For `gpt-4o-transcribe` and
133- `gpt-4o-mini-transcribe` , the only supported format is `json` . For
134- `gpt-4o-transcribe-diarize` , the supported formats are `json` , `text` , and
135- `diarized_json` , with `diarized_json` required to receive speaker annotations .
136-
137- stream : If set to true , the model response data will be streamed to the client as it is
138- generated using
139- [server - sent events ](https :// developer .mozilla .org / en - US / docs / Web / API / Server - sent_events / Using_server - sent_events #Event_stream_format).
140- See the
141- [Streaming section of the Speech - to - Text guide ](https :// platform .openai .com / docs / guides / speech - to - text ?lang = curl #streaming-transcriptions)
142- for more information .
143-
144- Note : Streaming is not supported for the `whisper-1` model and will be ignored .
145-
146- temperature : The sampling temperature , between 0 and 1. Higher values like 0.8 will make the
147- output more random , while lower values like 0.2 will make it more focused and
148- deterministic . If set to 0 , the model will use
149- [log probability ](https :// en .wikipedia .org / wiki / Log_probability ) to
150- automatically increase the temperature until certain thresholds are hit .
151-
152- timestamp_granularities : The timestamp granularities to populate for this transcription .
153- `response_format` must be set `verbose_json` to use timestamp granularities .
154- Either or both of these options are supported : `word` , or `segment` . Note : There
155- is no additional latency for segment timestamps , but generating word timestamps
156- incurs additional latency . This option is not available for
157- `gpt-4o-transcribe-diarize` .
158-
159- extra_headers : Send extra headers
160-
161- extra_query : Add additional query parameters to the request
162-
163- extra_body : Add additional JSON properties to the request
164- ) -> Transcription : ...
165-
166106 @overload
167107 def create (
168108 self ,
0 commit comments