cli/resources/audio/subresources/transcriptions/index.md +0 −965 deleted
File Deleted View Diff
1# Transcriptions
2
3## Create transcription
4
5`$ openai audio:transcriptions create`
6
7**post** `/audio/transcriptions`
8
9Transcribes audio into the input language.
10
11Returns a transcription object in `json`, `diarized_json`, or `verbose_json`
12format, or a stream of transcript events.
13
14### Parameters
15
16- `--file: string`
17
18 The audio file object (not file name) to transcribe, in one of these formats: flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.
19
20- `--model: string or AudioModel`
21
22 ID of the model to use. The options are `gpt-4o-transcribe`, `gpt-4o-mini-transcribe`, `gpt-4o-mini-transcribe-2025-12-15`, `whisper-1` (which is powered by our open source Whisper V2 model), and `gpt-4o-transcribe-diarize`.
23
24- `--chunking-strategy: optional "auto" or object { type, prefix_padding_ms, silence_duration_ms, threshold }`
25
26 Controls how the audio is cut into chunks. When set to `"auto"`, the server first normalizes loudness and then uses voice activity detection (VAD) to choose boundaries. `server_vad` object can be provided to tweak VAD detection parameters manually. If unset, the audio is transcribed as a single block. Required when using `gpt-4o-transcribe-diarize` for inputs longer than 30 seconds.
27
28- `--include: optional array of TranscriptionInclude`
29
30 Additional information to include in the transcription response.
31 `logprobs` will return the log probabilities of the tokens in the
32 response to understand the model's confidence in the transcription.
33 `logprobs` only works with response_format set to `json` and only with
34 the models `gpt-4o-transcribe`, `gpt-4o-mini-transcribe`, and `gpt-4o-mini-transcribe-2025-12-15`. This field is not supported when using `gpt-4o-transcribe-diarize`.
35
36- `--known-speaker-name: optional array of string`
37
38 Optional list of speaker names that correspond to the audio samples provided in `known_speaker_references[]`. Each entry should be a short identifier (for example `customer` or `agent`). Up to 4 speakers are supported.
39
40- `--known-speaker-reference: optional array of string`
41
42 Optional list of audio samples (as [data URLs](https://developer.mozilla.org/en-US/docs/Web/HTTP/Basics_of_HTTP/Data_URLs)) that contain known speaker references matching `known_speaker_names[]`. Each sample must be between 2 and 10 seconds, and can use any of the same input audio formats supported by `file`.
43
44- `--language: optional string`
45
46 The language of the input audio. Supplying the input language in [ISO-639-1](https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes) (e.g. `en`) format will improve accuracy and latency.
47
48- `--prompt: optional string`
49
50 An optional text to guide the model's style or continue a previous audio segment. The [prompt](https://platform.openai.com/docs/guides/speech-to-text#prompting) should match the audio language. This field is not supported when using `gpt-4o-transcribe-diarize`.
51
52- `--response-format: optional "json" or "text" or "srt" or 3 more`
53
54 The format of the output, in one of these options: `json`, `text`, `srt`, `verbose_json`, `vtt`, or `diarized_json`. For `gpt-4o-transcribe` and `gpt-4o-mini-transcribe`, the only supported format is `json`. For `gpt-4o-transcribe-diarize`, the supported formats are `json`, `text`, and `diarized_json`, with `diarized_json` required to receive speaker annotations.
55
56- `--temperature: optional number`
57
58 The sampling temperature, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. If set to 0, the model will use [log probability](https://en.wikipedia.org/wiki/Log_probability) to automatically increase the temperature until certain thresholds are hit.
59
60- `--timestamp-granularity: optional array of "word" or "segment"`
61
62 The timestamp granularities to populate for this transcription. `response_format` must be set `verbose_json` to use timestamp granularities. Either or both of these options are supported: `word`, or `segment`. Note: There is no additional latency for segment timestamps, but generating word timestamps incurs additional latency.
63 This option is not available for `gpt-4o-transcribe-diarize`.
64
65### Returns
66
67- `AudioTranscriptionNewResponse: Transcription or TranscriptionDiarized or TranscriptionVerbose`
68
69 Represents a transcription response returned by model, based on the provided input.
70
71 - `transcription: object { text, logprobs, usage }`
72
73 Represents a transcription response returned by model, based on the provided input.
74
75 - `text: string`
76
77 The transcribed text.
78
79 - `logprobs: optional array of object { token, bytes, logprob }`
80
81 The log probabilities of the tokens in the transcription. Only returned with the models `gpt-4o-transcribe` and `gpt-4o-mini-transcribe` if `logprobs` is added to the `include` array.
82
83 - `token: optional string`
84
85 The token in the transcription.
86
87 - `bytes: optional array of number`
88
89 The bytes of the token.
90
91 - `logprob: optional number`
92
93 The log probability of the token.
94
95 - `usage: optional object { input_tokens, output_tokens, total_tokens, 2 more } or object { seconds, type }`
96
97 Token usage statistics for the request.
98
99 - `tokens: object { input_tokens, output_tokens, total_tokens, 2 more }`
100
101 Usage statistics for models billed by token usage.
102
103 - `input_tokens: number`
104
105 Number of input tokens billed for this request.
106
107 - `output_tokens: number`
108
109 Number of output tokens generated.
110
111 - `total_tokens: number`
112
113 Total number of tokens used (input + output).
114
115 - `type: "tokens"`
116
117 The type of the usage object. Always `tokens` for this variant.
118
119 - `input_token_details: optional object { audio_tokens, text_tokens }`
120
121 Details about the input tokens billed for this request.
122
123 - `audio_tokens: optional number`
124
125 Number of audio tokens billed for this request.
126
127 - `text_tokens: optional number`
128
129 Number of text tokens billed for this request.
130
131 - `duration: object { seconds, type }`
132
133 Usage statistics for models billed by audio input duration.
134
135 - `seconds: number`
136
137 Duration of the input audio in seconds.
138
139 - `type: "duration"`
140
141 The type of the usage object. Always `duration` for this variant.
142
143 - `transcription_diarized: object { duration, segments, task, 2 more }`
144
145 Represents a diarized transcription response returned by the model, including the combined transcript and speaker-segment annotations.
146
147 - `duration: number`
148
149 Duration of the input audio in seconds.
150
151 - `segments: array of TranscriptionDiarizedSegment`
152
153 Segments of the transcript annotated with timestamps and speaker labels.
154
155 - `id: string`
156
157 Unique identifier for the segment.
158
159 - `end: number`
160
161 End timestamp of the segment in seconds.
162
163 - `speaker: string`
164
165 Speaker label for this segment. When known speakers are provided, the label matches `known_speaker_names[]`. Otherwise speakers are labeled sequentially using capital letters (`A`, `B`, ...).
166
167 - `start: number`
168
169 Start timestamp of the segment in seconds.
170
171 - `text: string`
172
173 Transcript text for this segment.
174
175 - `type: "transcript.text.segment"`
176
177 The type of the segment. Always `transcript.text.segment`.
178
179 - `task: "transcribe"`
180
181 The type of task that was run. Always `transcribe`.
182
183 - `text: string`
184
185 The concatenated transcript text for the entire audio input.
186
187 - `usage: optional object { input_tokens, output_tokens, total_tokens, 2 more } or object { seconds, type }`
188
189 Token or duration usage statistics for the request.
190
191 - `tokens: object { input_tokens, output_tokens, total_tokens, 2 more }`
192
193 Usage statistics for models billed by token usage.
194
195 - `input_tokens: number`
196
197 Number of input tokens billed for this request.
198
199 - `output_tokens: number`
200
201 Number of output tokens generated.
202
203 - `total_tokens: number`
204
205 Total number of tokens used (input + output).
206
207 - `type: "tokens"`
208
209 The type of the usage object. Always `tokens` for this variant.
210
211 - `input_token_details: optional object { audio_tokens, text_tokens }`
212
213 Details about the input tokens billed for this request.
214
215 - `audio_tokens: optional number`
216
217 Number of audio tokens billed for this request.
218
219 - `text_tokens: optional number`
220
221 Number of text tokens billed for this request.
222
223 - `duration: object { seconds, type }`
224
225 Usage statistics for models billed by audio input duration.
226
227 - `seconds: number`
228
229 Duration of the input audio in seconds.
230
231 - `type: "duration"`
232
233 The type of the usage object. Always `duration` for this variant.
234
235 - `transcription_verbose: object { duration, language, text, 3 more }`
236
237 Represents a verbose json transcription response returned by model, based on the provided input.
238
239 - `duration: number`
240
241 The duration of the input audio.
242
243 - `language: string`
244
245 The language of the input audio.
246
247 - `text: string`
248
249 The transcribed text.
250
251 - `segments: optional array of TranscriptionSegment`
252
253 Segments of the transcribed text and their corresponding details.
254
255 - `id: number`
256
257 Unique identifier of the segment.
258
259 - `avg_logprob: number`
260
261 Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.
262
263 - `compression_ratio: number`
264
265 Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.
266
267 - `end: number`
268
269 End time of the segment in seconds.
270
271 - `no_speech_prob: number`
272
273 Probability of no speech in the segment. If the value is higher than 1.0 and the `avg_logprob` is below -1, consider this segment silent.
274
275 - `seek: number`
276
277 Seek offset of the segment.
278
279 - `start: number`
280
281 Start time of the segment in seconds.
282
283 - `temperature: number`
284
285 Temperature parameter used for generating the segment.
286
287 - `text: string`
288
289 Text content of the segment.
290
291 - `tokens: array of number`
292
293 Array of token IDs for the text content.
294
295 - `usage: optional object { seconds, type }`
296
297 Usage statistics for models billed by audio input duration.
298
299 - `seconds: number`
300
301 Duration of the input audio in seconds.
302
303 - `type: "duration"`
304
305 The type of the usage object. Always `duration` for this variant.
306
307 - `words: optional array of TranscriptionWord`
308
309 Extracted words and their corresponding timestamps.
310
311 - `end: number`
312
313 End time of the word in seconds.
314
315 - `start: number`
316
317 Start time of the word in seconds.
318
319 - `word: string`
320
321 The text content of the word.
322
323### Example
324
325```cli
326openai audio:transcriptions create \
327 --api-key 'My API Key' \
328 --file 'Example data' \
329 --model gpt-4o-transcribe
330```
331
332#### Response
333
334```json
335{
336 "text": "text",
337 "logprobs": [
338 {
339 "token": "token",
340 "bytes": [
341 0
342 ],
343 "logprob": 0
344 }
345 ],
346 "usage": {
347 "input_tokens": 0,
348 "output_tokens": 0,
349 "total_tokens": 0,
350 "type": "tokens",
351 "input_token_details": {
352 "audio_tokens": 0,
353 "text_tokens": 0
354 }
355 }
356}
357```
358
359## Domain Types
360
361### Transcription
362
363- `transcription: object { text, logprobs, usage }`
364
365 Represents a transcription response returned by model, based on the provided input.
366
367 - `text: string`
368
369 The transcribed text.
370
371 - `logprobs: optional array of object { token, bytes, logprob }`
372
373 The log probabilities of the tokens in the transcription. Only returned with the models `gpt-4o-transcribe` and `gpt-4o-mini-transcribe` if `logprobs` is added to the `include` array.
374
375 - `token: optional string`
376
377 The token in the transcription.
378
379 - `bytes: optional array of number`
380
381 The bytes of the token.
382
383 - `logprob: optional number`
384
385 The log probability of the token.
386
387 - `usage: optional object { input_tokens, output_tokens, total_tokens, 2 more } or object { seconds, type }`
388
389 Token usage statistics for the request.
390
391 - `tokens: object { input_tokens, output_tokens, total_tokens, 2 more }`
392
393 Usage statistics for models billed by token usage.
394
395 - `input_tokens: number`
396
397 Number of input tokens billed for this request.
398
399 - `output_tokens: number`
400
401 Number of output tokens generated.
402
403 - `total_tokens: number`
404
405 Total number of tokens used (input + output).
406
407 - `type: "tokens"`
408
409 The type of the usage object. Always `tokens` for this variant.
410
411 - `input_token_details: optional object { audio_tokens, text_tokens }`
412
413 Details about the input tokens billed for this request.
414
415 - `audio_tokens: optional number`
416
417 Number of audio tokens billed for this request.
418
419 - `text_tokens: optional number`
420
421 Number of text tokens billed for this request.
422
423 - `duration: object { seconds, type }`
424
425 Usage statistics for models billed by audio input duration.
426
427 - `seconds: number`
428
429 Duration of the input audio in seconds.
430
431 - `type: "duration"`
432
433 The type of the usage object. Always `duration` for this variant.
434
435### Transcription Diarized
436
437- `transcription_diarized: object { duration, segments, task, 2 more }`
438
439 Represents a diarized transcription response returned by the model, including the combined transcript and speaker-segment annotations.
440
441 - `duration: number`
442
443 Duration of the input audio in seconds.
444
445 - `segments: array of TranscriptionDiarizedSegment`
446
447 Segments of the transcript annotated with timestamps and speaker labels.
448
449 - `id: string`
450
451 Unique identifier for the segment.
452
453 - `end: number`
454
455 End timestamp of the segment in seconds.
456
457 - `speaker: string`
458
459 Speaker label for this segment. When known speakers are provided, the label matches `known_speaker_names[]`. Otherwise speakers are labeled sequentially using capital letters (`A`, `B`, ...).
460
461 - `start: number`
462
463 Start timestamp of the segment in seconds.
464
465 - `text: string`
466
467 Transcript text for this segment.
468
469 - `type: "transcript.text.segment"`
470
471 The type of the segment. Always `transcript.text.segment`.
472
473 - `task: "transcribe"`
474
475 The type of task that was run. Always `transcribe`.
476
477 - `text: string`
478
479 The concatenated transcript text for the entire audio input.
480
481 - `usage: optional object { input_tokens, output_tokens, total_tokens, 2 more } or object { seconds, type }`
482
483 Token or duration usage statistics for the request.
484
485 - `tokens: object { input_tokens, output_tokens, total_tokens, 2 more }`
486
487 Usage statistics for models billed by token usage.
488
489 - `input_tokens: number`
490
491 Number of input tokens billed for this request.
492
493 - `output_tokens: number`
494
495 Number of output tokens generated.
496
497 - `total_tokens: number`
498
499 Total number of tokens used (input + output).
500
501 - `type: "tokens"`
502
503 The type of the usage object. Always `tokens` for this variant.
504
505 - `input_token_details: optional object { audio_tokens, text_tokens }`
506
507 Details about the input tokens billed for this request.
508
509 - `audio_tokens: optional number`
510
511 Number of audio tokens billed for this request.
512
513 - `text_tokens: optional number`
514
515 Number of text tokens billed for this request.
516
517 - `duration: object { seconds, type }`
518
519 Usage statistics for models billed by audio input duration.
520
521 - `seconds: number`
522
523 Duration of the input audio in seconds.
524
525 - `type: "duration"`
526
527 The type of the usage object. Always `duration` for this variant.
528
529### Transcription Diarized Segment
530
531- `transcription_diarized_segment: object { id, end, speaker, 3 more }`
532
533 A segment of diarized transcript text with speaker metadata.
534
535 - `id: string`
536
537 Unique identifier for the segment.
538
539 - `end: number`
540
541 End timestamp of the segment in seconds.
542
543 - `speaker: string`
544
545 Speaker label for this segment. When known speakers are provided, the label matches `known_speaker_names[]`. Otherwise speakers are labeled sequentially using capital letters (`A`, `B`, ...).
546
547 - `start: number`
548
549 Start timestamp of the segment in seconds.
550
551 - `text: string`
552
553 Transcript text for this segment.
554
555 - `type: "transcript.text.segment"`
556
557 The type of the segment. Always `transcript.text.segment`.
558
559### Transcription Include
560
561- `transcription_include: "logprobs"`
562
563 - `"logprobs"`
564
565### Transcription Segment
566
567- `transcription_segment: object { id, avg_logprob, compression_ratio, 7 more }`
568
569 - `id: number`
570
571 Unique identifier of the segment.
572
573 - `avg_logprob: number`
574
575 Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.
576
577 - `compression_ratio: number`
578
579 Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.
580
581 - `end: number`
582
583 End time of the segment in seconds.
584
585 - `no_speech_prob: number`
586
587 Probability of no speech in the segment. If the value is higher than 1.0 and the `avg_logprob` is below -1, consider this segment silent.
588
589 - `seek: number`
590
591 Seek offset of the segment.
592
593 - `start: number`
594
595 Start time of the segment in seconds.
596
597 - `temperature: number`
598
599 Temperature parameter used for generating the segment.
600
601 - `text: string`
602
603 Text content of the segment.
604
605 - `tokens: array of number`
606
607 Array of token IDs for the text content.
608
609### Transcription Stream Event
610
611- `transcription_stream_event: TranscriptionTextSegmentEvent or TranscriptionTextDeltaEvent or TranscriptionTextDoneEvent`
612
613 Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with `stream` set to `true` and `response_format` set to `diarized_json`.
614
615 - `transcription_text_segment_event: object { id, end, speaker, 3 more }`
616
617 Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with `stream` set to `true` and `response_format` set to `diarized_json`.
618
619 - `id: string`
620
621 Unique identifier for the segment.
622
623 - `end: number`
624
625 End timestamp of the segment in seconds.
626
627 - `speaker: string`
628
629 Speaker label for this segment.
630
631 - `start: number`
632
633 Start timestamp of the segment in seconds.
634
635 - `text: string`
636
637 Transcript text for this segment.
638
639 - `type: "transcript.text.segment"`
640
641 The type of the event. Always `transcript.text.segment`.
642
643 - `transcription_text_delta_event: object { delta, type, logprobs, segment_id }`
644
645 Emitted when there is an additional text delta. This is also the first event emitted when the transcription starts. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `Stream` parameter set to `true`.
646
647 - `delta: string`
648
649 The text delta that was additionally transcribed.
650
651 - `type: "transcript.text.delta"`
652
653 The type of the event. Always `transcript.text.delta`.
654
655 - `logprobs: optional array of object { token, bytes, logprob }`
656
657 The log probabilities of the delta. Only included if you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `include[]` parameter set to `logprobs`.
658
659 - `token: optional string`
660
661 The token that was used to generate the log probability.
662
663 - `bytes: optional array of number`
664
665 The bytes that were used to generate the log probability.
666
667 - `logprob: optional number`
668
669 The log probability of the token.
670
671 - `segment_id: optional string`
672
673 Identifier of the diarized segment that this delta belongs to. Only present when using `gpt-4o-transcribe-diarize`.
674
675 - `transcription_text_done_event: object { text, type, logprobs, usage }`
676
677 Emitted when the transcription is complete. Contains the complete transcription text. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `Stream` parameter set to `true`.
678
679 - `text: string`
680
681 The text that was transcribed.
682
683 - `type: "transcript.text.done"`
684
685 The type of the event. Always `transcript.text.done`.
686
687 - `logprobs: optional array of object { token, bytes, logprob }`
688
689 The log probabilities of the individual tokens in the transcription. Only included if you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `include[]` parameter set to `logprobs`.
690
691 - `token: optional string`
692
693 The token that was used to generate the log probability.
694
695 - `bytes: optional array of number`
696
697 The bytes that were used to generate the log probability.
698
699 - `logprob: optional number`
700
701 The log probability of the token.
702
703 - `usage: optional object { input_tokens, output_tokens, total_tokens, 2 more }`
704
705 Usage statistics for models billed by token usage.
706
707 - `input_tokens: number`
708
709 Number of input tokens billed for this request.
710
711 - `output_tokens: number`
712
713 Number of output tokens generated.
714
715 - `total_tokens: number`
716
717 Total number of tokens used (input + output).
718
719 - `type: "tokens"`
720
721 The type of the usage object. Always `tokens` for this variant.
722
723 - `input_token_details: optional object { audio_tokens, text_tokens }`
724
725 Details about the input tokens billed for this request.
726
727 - `audio_tokens: optional number`
728
729 Number of audio tokens billed for this request.
730
731 - `text_tokens: optional number`
732
733 Number of text tokens billed for this request.
734
735### Transcription Text Delta Event
736
737- `transcription_text_delta_event: object { delta, type, logprobs, segment_id }`
738
739 Emitted when there is an additional text delta. This is also the first event emitted when the transcription starts. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `Stream` parameter set to `true`.
740
741 - `delta: string`
742
743 The text delta that was additionally transcribed.
744
745 - `type: "transcript.text.delta"`
746
747 The type of the event. Always `transcript.text.delta`.
748
749 - `logprobs: optional array of object { token, bytes, logprob }`
750
751 The log probabilities of the delta. Only included if you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `include[]` parameter set to `logprobs`.
752
753 - `token: optional string`
754
755 The token that was used to generate the log probability.
756
757 - `bytes: optional array of number`
758
759 The bytes that were used to generate the log probability.
760
761 - `logprob: optional number`
762
763 The log probability of the token.
764
765 - `segment_id: optional string`
766
767 Identifier of the diarized segment that this delta belongs to. Only present when using `gpt-4o-transcribe-diarize`.
768
769### Transcription Text Done Event
770
771- `transcription_text_done_event: object { text, type, logprobs, usage }`
772
773 Emitted when the transcription is complete. Contains the complete transcription text. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `Stream` parameter set to `true`.
774
775 - `text: string`
776
777 The text that was transcribed.
778
779 - `type: "transcript.text.done"`
780
781 The type of the event. Always `transcript.text.done`.
782
783 - `logprobs: optional array of object { token, bytes, logprob }`
784
785 The log probabilities of the individual tokens in the transcription. Only included if you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with the `include[]` parameter set to `logprobs`.
786
787 - `token: optional string`
788
789 The token that was used to generate the log probability.
790
791 - `bytes: optional array of number`
792
793 The bytes that were used to generate the log probability.
794
795 - `logprob: optional number`
796
797 The log probability of the token.
798
799 - `usage: optional object { input_tokens, output_tokens, total_tokens, 2 more }`
800
801 Usage statistics for models billed by token usage.
802
803 - `input_tokens: number`
804
805 Number of input tokens billed for this request.
806
807 - `output_tokens: number`
808
809 Number of output tokens generated.
810
811 - `total_tokens: number`
812
813 Total number of tokens used (input + output).
814
815 - `type: "tokens"`
816
817 The type of the usage object. Always `tokens` for this variant.
818
819 - `input_token_details: optional object { audio_tokens, text_tokens }`
820
821 Details about the input tokens billed for this request.
822
823 - `audio_tokens: optional number`
824
825 Number of audio tokens billed for this request.
826
827 - `text_tokens: optional number`
828
829 Number of text tokens billed for this request.
830
831### Transcription Text Segment Event
832
833- `transcription_text_segment_event: object { id, end, speaker, 3 more }`
834
835 Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription) with `stream` set to `true` and `response_format` set to `diarized_json`.
836
837 - `id: string`
838
839 Unique identifier for the segment.
840
841 - `end: number`
842
843 End timestamp of the segment in seconds.
844
845 - `speaker: string`
846
847 Speaker label for this segment.
848
849 - `start: number`
850
851 Start timestamp of the segment in seconds.
852
853 - `text: string`
854
855 Transcript text for this segment.
856
857 - `type: "transcript.text.segment"`
858
859 The type of the event. Always `transcript.text.segment`.
860
861### Transcription Verbose
862
863- `transcription_verbose: object { duration, language, text, 3 more }`
864
865 Represents a verbose json transcription response returned by model, based on the provided input.
866
867 - `duration: number`
868
869 The duration of the input audio.
870
871 - `language: string`
872
873 The language of the input audio.
874
875 - `text: string`
876
877 The transcribed text.
878
879 - `segments: optional array of TranscriptionSegment`
880
881 Segments of the transcribed text and their corresponding details.
882
883 - `id: number`
884
885 Unique identifier of the segment.
886
887 - `avg_logprob: number`
888
889 Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.
890
891 - `compression_ratio: number`
892
893 Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.
894
895 - `end: number`
896
897 End time of the segment in seconds.
898
899 - `no_speech_prob: number`
900
901 Probability of no speech in the segment. If the value is higher than 1.0 and the `avg_logprob` is below -1, consider this segment silent.
902
903 - `seek: number`
904
905 Seek offset of the segment.
906
907 - `start: number`
908
909 Start time of the segment in seconds.
910
911 - `temperature: number`
912
913 Temperature parameter used for generating the segment.
914
915 - `text: string`
916
917 Text content of the segment.
918
919 - `tokens: array of number`
920
921 Array of token IDs for the text content.
922
923 - `usage: optional object { seconds, type }`
924
925 Usage statistics for models billed by audio input duration.
926
927 - `seconds: number`
928
929 Duration of the input audio in seconds.
930
931 - `type: "duration"`
932
933 The type of the usage object. Always `duration` for this variant.
934
935 - `words: optional array of TranscriptionWord`
936
937 Extracted words and their corresponding timestamps.
938
939 - `end: number`
940
941 End time of the word in seconds.
942
943 - `start: number`
944
945 Start time of the word in seconds.
946
947 - `word: string`
948
949 The text content of the word.
950
951### Transcription Word
952
953- `transcription_word: object { end, start, word }`
954
955 - `end: number`
956
957 End time of the word in seconds.
958
959 - `start: number`
960
961 Start time of the word in seconds.
962
963 - `word: string`
964
965 The text content of the word.