Transcriptions

Create transcription

client.Audio.Transcriptions.New(ctx, body) (*AudioTranscriptionNewResponseUnion, error)

post /audio/transcriptions

Transcribes audio into the input language.

Returns a transcription object in json, diarized_json, or verbose_json format, or a stream of transcript events.

Parameters

body AudioTranscriptionNewParams
- File param.Field[Reader]
  
  The audio file object (not file name) to transcribe, in one of these formats: flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.
- Model param.Field[AudioModel]
  
  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.
  - string
  - type AudioModel string
    - const AudioModelWhisper1 AudioModel = "whisper-1"
    - const AudioModelGPT4oTranscribe AudioModel = "gpt-4o-transcribe"
    - const AudioModelGPT4oMiniTranscribe AudioModel = "gpt-4o-mini-transcribe"
    - const AudioModelGPT4oMiniTranscribe2025_12_15 AudioModel = "gpt-4o-mini-transcribe-2025-12-15"
    - const AudioModelGPT4oTranscribeDiarize AudioModel = "gpt-4o-transcribe-diarize"
- ChunkingStrategy param.Field[AudioTranscriptionNewParamsChunkingStrategyUnion]
  
  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.
  - Auto
    - const AutoAuto Auto = "auto"
  - AudioTranscriptionNewParamsChunkingStrategyVadConfig
    - Type string
      
      Must be set to server_vad to enable manual chunking using server side VAD.
      - const AudioTranscriptionNewParamsChunkingStrategyVadConfigTypeServerVad AudioTranscriptionNewParamsChunkingStrategyVadConfigType = "server_vad"
    - PrefixPaddingMs int64
      
      Amount of audio to include before the VAD detected speech (in milliseconds).
    - SilenceDurationMs int64
      
      Duration of silence to detect speech stop (in milliseconds). With shorter values the model will respond more quickly, but may jump in on short pauses from the user.
    - Threshold float64
      
      Sensitivity threshold (0.0 to 1.0) for voice activity detection. A higher threshold will require louder audio to activate the model, and thus might perform better in noisy environments.
- Include param.Field[[]TranscriptionInclude]
  
  Additional information to include in the transcription response. logprobs will return the log probabilities of the tokens in the response to understand the model's confidence in the transcription. logprobs only works with response_format set to json and only with 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.
  - const TranscriptionIncludeLogprobs TranscriptionInclude = "logprobs"
- KnownSpeakerNames param.Field[[]string]
  
  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.
- KnownSpeakerReferences param.Field[[]string]
  
  Optional list of audio samples (as 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.
- Language param.Field[string]
  
  The language of the input audio. Supplying the input language in ISO-639-1 (e.g. en) format will improve accuracy and latency.
- Prompt param.Field[string]
  
  An optional text to guide the model's style or continue a previous audio segment. The prompt should match the audio language. This field is not supported when using gpt-4o-transcribe-diarize.
- ResponseFormat param.Field[AudioResponseFormat]
  
  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.
- ``
- Temperature param.Field[float64]
  
  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 to automatically increase the temperature until certain thresholds are hit.
- TimestampGranularities param.Field[[]string]
  
  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. This option is not available for gpt-4o-transcribe-diarize.
  - const AudioTranscriptionNewParamsTimestampGranularityWord AudioTranscriptionNewParamsTimestampGranularity = "word"
  - const AudioTranscriptionNewParamsTimestampGranularitySegment AudioTranscriptionNewParamsTimestampGranularity = "segment"

Returns

type AudioTranscriptionNewResponseUnion interface{…}

Represents a transcription response returned by model, based on the provided input.
- type Transcription struct{…}
  
  Represents a transcription response returned by model, based on the provided input.
  - Text string
    
    The transcribed text.
  - Logprobs []TranscriptionLogprob
    
    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.
    - Token string
      
      The token in the transcription.
    - Bytes []float64
      
      The bytes of the token.
    - Logprob float64
      
      The log probability of the token.
  - Usage TranscriptionUsageUnion
    
    Token usage statistics for the request.
    - type TranscriptionUsageTokens struct{…}
      
      Usage statistics for models billed by token usage.
      - InputTokens int64
        
        Number of input tokens billed for this request.
      - OutputTokens int64
        
        Number of output tokens generated.
      - TotalTokens int64
        
        Total number of tokens used (input + output).
      - Type Tokens
        
        The type of the usage object. Always tokens for this variant.
        
        const TokensTokens Tokens = "tokens"
      - InputTokenDetails TranscriptionUsageTokensInputTokenDetails
        
        Details about the input tokens billed for this request.
        
        AudioTokens int64
        
        Number of audio tokens billed for this request.
        
        TextTokens int64
        
        Number of text tokens billed for this request.
    - type TranscriptionUsageDuration struct{…}
      
      Usage statistics for models billed by audio input duration.
      - Seconds float64
        
        Duration of the input audio in seconds.
      - Type Duration
        
        The type of the usage object. Always duration for this variant.
        
        const DurationDuration Duration = "duration"
- type TranscriptionVerbose struct{…}
  
  Represents a verbose json transcription response returned by model, based on the provided input.
  - Duration float64
    
    The duration of the input audio.
  - Language string
    
    The language of the input audio.
  - Text string
    
    The transcribed text.
  - Segments []TranscriptionSegment
    
    Segments of the transcribed text and their corresponding details.
    - ID int64
      
      Unique identifier of the segment.
    - AvgLogprob float64
      
      Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.
    - CompressionRatio float64
      
      Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.
    - End float64
      
      End time of the segment in seconds.
    - NoSpeechProb float64
      
      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.
    - Seek int64
      
      Seek offset of the segment.
    - Start float64
      
      Start time of the segment in seconds.
    - Temperature float64
      
      Temperature parameter used for generating the segment.
    - Text string
      
      Text content of the segment.
    - Tokens []int64
      
      Array of token IDs for the text content.
  - Usage TranscriptionVerboseUsage
    
    Usage statistics for models billed by audio input duration.
    - Seconds float64
      
      Duration of the input audio in seconds.
    - Type Duration
      
      The type of the usage object. Always duration for this variant.
      - const DurationDuration Duration = "duration"
  - Words []TranscriptionWord
    
    Extracted words and their corresponding timestamps.
    - End float64
      
      End time of the word in seconds.
    - Start float64
      
      Start time of the word in seconds.
    - Word string
      
      The text content of the word.

Example

package main

import (
  "bytes"
  "context"
  "fmt"
  "io"

  "github.com/openai/openai-go"
  "github.com/openai/openai-go/option"
)

func main() {
  client := openai.NewClient(
    option.WithAPIKey("My API Key"),
  )
  transcription, err := client.Audio.Transcriptions.New(context.TODO(), openai.AudioTranscriptionNewParams{
    File: io.Reader(bytes.NewBuffer([]byte("Example data"))),
    Model: openai.AudioModelGPT4oTranscribe,
  })
  if err != nil {
    panic(err.Error())
  }
  fmt.Printf("%+v\n", transcription)
}

Response

{
  "text": "text",
  "logprobs": [
    {
      "token": "token",
      "bytes": [
        0
      ],
      "logprob": 0
    }
  ],
  "usage": {
    "input_tokens": 0,
    "output_tokens": 0,
    "total_tokens": 0,
    "type": "tokens",
    "input_token_details": {
      "audio_tokens": 0,
      "text_tokens": 0
    }
  }
}

Domain Types

Transcription

type Transcription struct{…}

Represents a transcription response returned by model, based on the provided input.
- Text string
  
  The transcribed text.
- Logprobs []TranscriptionLogprob
  
  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.
  - Token string
    
    The token in the transcription.
  - Bytes []float64
    
    The bytes of the token.
  - Logprob float64
    
    The log probability of the token.
- Usage TranscriptionUsageUnion
  
  Token usage statistics for the request.
  - type TranscriptionUsageTokens struct{…}
    
    Usage statistics for models billed by token usage.
    - InputTokens int64
      
      Number of input tokens billed for this request.
    - OutputTokens int64
      
      Number of output tokens generated.
    - TotalTokens int64
      
      Total number of tokens used (input + output).
    - Type Tokens
      
      The type of the usage object. Always tokens for this variant.
      - const TokensTokens Tokens = "tokens"
    - InputTokenDetails TranscriptionUsageTokensInputTokenDetails
      
      Details about the input tokens billed for this request.
      - AudioTokens int64
        
        Number of audio tokens billed for this request.
      - TextTokens int64
        
        Number of text tokens billed for this request.
  - type TranscriptionUsageDuration struct{…}
    
    Usage statistics for models billed by audio input duration.
    - Seconds float64
      
      Duration of the input audio in seconds.
    - Type Duration
      
      The type of the usage object. Always duration for this variant.
      - const DurationDuration Duration = "duration"

Transcription Diarized

type TranscriptionDiarized struct{…}

Represents a diarized transcription response returned by the model, including the combined transcript and speaker-segment annotations.
- Duration float64
  
  Duration of the input audio in seconds.
- Segments []TranscriptionDiarizedSegment
  
  Segments of the transcript annotated with timestamps and speaker labels.
  - ID string
    
    Unique identifier for the segment.
  - End float64
    
    End timestamp of the segment in seconds.
  - Speaker string
    
    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, ...).
  - Start float64
    
    Start timestamp of the segment in seconds.
  - Text string
    
    Transcript text for this segment.
  - Type TranscriptTextSegment
    
    The type of the segment. Always transcript.text.segment.
    - const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"
- Task Transcribe
  
  The type of task that was run. Always transcribe.
  - const TranscribeTranscribe Transcribe = "transcribe"
- Text string
  
  The concatenated transcript text for the entire audio input.
- Usage TranscriptionDiarizedUsageUnion
  
  Token or duration usage statistics for the request.
  - TranscriptionDiarizedUsageTokens
    - InputTokens int64
      
      Number of input tokens billed for this request.
    - OutputTokens int64
      
      Number of output tokens generated.
    - TotalTokens int64
      
      Total number of tokens used (input + output).
    - Type Tokens
      
      The type of the usage object. Always tokens for this variant.
      - const TokensTokens Tokens = "tokens"
    - InputTokenDetails TranscriptionDiarizedUsageTokensInputTokenDetails
      
      Details about the input tokens billed for this request.
      - AudioTokens int64
        
        Number of audio tokens billed for this request.
      - TextTokens int64
        
        Number of text tokens billed for this request.
  - TranscriptionDiarizedUsageDuration
    - Seconds float64
      
      Duration of the input audio in seconds.
    - Type Duration
      
      The type of the usage object. Always duration for this variant.
      - const DurationDuration Duration = "duration"

Transcription Diarized Segment

type TranscriptionDiarizedSegment struct{…}

A segment of diarized transcript text with speaker metadata.
- ID string
  
  Unique identifier for the segment.
- End float64
  
  End timestamp of the segment in seconds.
- Speaker string
  
  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, ...).
- Start float64
  
  Start timestamp of the segment in seconds.
- Text string
  
  Transcript text for this segment.
- Type TranscriptTextSegment
  
  The type of the segment. Always transcript.text.segment.
  - const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"

Transcription Include

type TranscriptionInclude string
- const TranscriptionIncludeLogprobs TranscriptionInclude = "logprobs"

Transcription Segment

type TranscriptionSegment struct{…}
- ID int64
  
  Unique identifier of the segment.
- AvgLogprob float64
  
  Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.
- CompressionRatio float64
  
  Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.
- End float64
  
  End time of the segment in seconds.
- NoSpeechProb float64
  
  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.
- Seek int64
  
  Seek offset of the segment.
- Start float64
  
  Start time of the segment in seconds.
- Temperature float64
  
  Temperature parameter used for generating the segment.
- Text string
  
  Text content of the segment.
- Tokens []int64
  
  Array of token IDs for the text content.

Transcription Stream Event

type TranscriptionStreamEventUnion interface{…}

Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you create a transcription with stream set to true and response_format set to diarized_json.
- type TranscriptionTextSegmentEvent struct{…}
  
  Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you create a transcription with stream set to true and response_format set to diarized_json.
  - ID string
    
    Unique identifier for the segment.
  - End float64
    
    End timestamp of the segment in seconds.
  - Speaker string
    
    Speaker label for this segment.
  - Start float64
    
    Start timestamp of the segment in seconds.
  - Text string
    
    Transcript text for this segment.
  - Type TranscriptTextSegment
    
    The type of the event. Always transcript.text.segment.
    - const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"
- type TranscriptionTextDeltaEvent struct{…}
  
  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 with the Stream parameter set to true.
  - Delta string
    
    The text delta that was additionally transcribed.
  - Type TranscriptTextDelta
    
    The type of the event. Always transcript.text.delta.
    - const TranscriptTextDeltaTranscriptTextDelta TranscriptTextDelta = "transcript.text.delta"
  - Logprobs []TranscriptionTextDeltaEventLogprob
    
    The log probabilities of the delta. Only included if you create a transcription with the include[] parameter set to logprobs.
    - Token string
      
      The token that was used to generate the log probability.
    - Bytes []int64
      
      The bytes that were used to generate the log probability.
    - Logprob float64
      
      The log probability of the token.
  - SegmentID string
    
    Identifier of the diarized segment that this delta belongs to. Only present when using gpt-4o-transcribe-diarize.
- type TranscriptionTextDoneEvent struct{…}
  
  Emitted when the transcription is complete. Contains the complete transcription text. Only emitted when you create a transcription with the Stream parameter set to true.
  - Text string
    
    The text that was transcribed.
  - Type TranscriptTextDone
    
    The type of the event. Always transcript.text.done.
    - const TranscriptTextDoneTranscriptTextDone TranscriptTextDone = "transcript.text.done"
  - Logprobs []TranscriptionTextDoneEventLogprob
    
    The log probabilities of the individual tokens in the transcription. Only included if you create a transcription with the include[] parameter set to logprobs.
    - Token string
      
      The token that was used to generate the log probability.
    - Bytes []int64
      
      The bytes that were used to generate the log probability.
    - Logprob float64
      
      The log probability of the token.
  - Usage TranscriptionTextDoneEventUsage
    
    Usage statistics for models billed by token usage.
    - InputTokens int64
      
      Number of input tokens billed for this request.
    - OutputTokens int64
      
      Number of output tokens generated.
    - TotalTokens int64
      
      Total number of tokens used (input + output).
    - Type Tokens
      
      The type of the usage object. Always tokens for this variant.
      - const TokensTokens Tokens = "tokens"
    - InputTokenDetails TranscriptionTextDoneEventUsageInputTokenDetails
      
      Details about the input tokens billed for this request.
      - AudioTokens int64
        
        Number of audio tokens billed for this request.
      - TextTokens int64
        
        Number of text tokens billed for this request.

Transcription Text Delta Event

type TranscriptionTextDeltaEvent struct{…}

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 with the Stream parameter set to true.
- Delta string
  
  The text delta that was additionally transcribed.
- Type TranscriptTextDelta
  
  The type of the event. Always transcript.text.delta.
  - const TranscriptTextDeltaTranscriptTextDelta TranscriptTextDelta = "transcript.text.delta"
- Logprobs []TranscriptionTextDeltaEventLogprob
  
  The log probabilities of the delta. Only included if you create a transcription with the include[] parameter set to logprobs.
  - Token string
    
    The token that was used to generate the log probability.
  - Bytes []int64
    
    The bytes that were used to generate the log probability.
  - Logprob float64
    
    The log probability of the token.
- SegmentID string
  
  Identifier of the diarized segment that this delta belongs to. Only present when using gpt-4o-transcribe-diarize.

Transcription Text Done Event

type TranscriptionTextDoneEvent struct{…}

Emitted when the transcription is complete. Contains the complete transcription text. Only emitted when you create a transcription with the Stream parameter set to true.
- Text string
  
  The text that was transcribed.
- Type TranscriptTextDone
  
  The type of the event. Always transcript.text.done.
  - const TranscriptTextDoneTranscriptTextDone TranscriptTextDone = "transcript.text.done"
- Logprobs []TranscriptionTextDoneEventLogprob
  
  The log probabilities of the individual tokens in the transcription. Only included if you create a transcription with the include[] parameter set to logprobs.
  - Token string
    
    The token that was used to generate the log probability.
  - Bytes []int64
    
    The bytes that were used to generate the log probability.
  - Logprob float64
    
    The log probability of the token.
- Usage TranscriptionTextDoneEventUsage
  
  Usage statistics for models billed by token usage.
  - InputTokens int64
    
    Number of input tokens billed for this request.
  - OutputTokens int64
    
    Number of output tokens generated.
  - TotalTokens int64
    
    Total number of tokens used (input + output).
  - Type Tokens
    
    The type of the usage object. Always tokens for this variant.
    - const TokensTokens Tokens = "tokens"
  - InputTokenDetails TranscriptionTextDoneEventUsageInputTokenDetails
    
    Details about the input tokens billed for this request.
    - AudioTokens int64
      
      Number of audio tokens billed for this request.
    - TextTokens int64
      
      Number of text tokens billed for this request.

Transcription Text Segment Event

type TranscriptionTextSegmentEvent struct{…}

Emitted when a diarized transcription returns a completed segment with speaker information. Only emitted when you create a transcription with stream set to true and response_format set to diarized_json.
- ID string
  
  Unique identifier for the segment.
- End float64
  
  End timestamp of the segment in seconds.
- Speaker string
  
  Speaker label for this segment.
- Start float64
  
  Start timestamp of the segment in seconds.
- Text string
  
  Transcript text for this segment.
- Type TranscriptTextSegment
  
  The type of the event. Always transcript.text.segment.
  - const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"

Transcription Verbose

type TranscriptionVerbose struct{…}

Represents a verbose json transcription response returned by model, based on the provided input.
- Duration float64
  
  The duration of the input audio.
- Language string
  
  The language of the input audio.
- Text string
  
  The transcribed text.
- Segments []TranscriptionSegment
  
  Segments of the transcribed text and their corresponding details.
  - ID int64
    
    Unique identifier of the segment.
  - AvgLogprob float64
    
    Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.
  - CompressionRatio float64
    
    Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.
  - End float64
    
    End time of the segment in seconds.
  - NoSpeechProb float64
    
    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.
  - Seek int64
    
    Seek offset of the segment.
  - Start float64
    
    Start time of the segment in seconds.
  - Temperature float64
    
    Temperature parameter used for generating the segment.
  - Text string
    
    Text content of the segment.
  - Tokens []int64
    
    Array of token IDs for the text content.
- Usage TranscriptionVerboseUsage
  
  Usage statistics for models billed by audio input duration.
  - Seconds float64
    
    Duration of the input audio in seconds.
  - Type Duration
    
    The type of the usage object. Always duration for this variant.
    - const DurationDuration Duration = "duration"
- Words []TranscriptionWord
  
  Extracted words and their corresponding timestamps.
  - End float64
    
    End time of the word in seconds.
  - Start float64
    
    Start time of the word in seconds.
  - Word string
    
    The text content of the word.

Transcription Word

type TranscriptionWord struct{…}
- End float64
  
  End time of the word in seconds.
- Start float64
  
  Start time of the word in seconds.
- Word string
  
  The text content of the word.

go/resources/audio/subresources/transcriptions/index.md +981 −0 created

1# Transcriptions

3## Create transcription

5`client.Audio.Transcriptions.New(ctx, body) (*AudioTranscriptionNewResponseUnion, error)`

7**post** `/audio/transcriptions`

9Transcribes audio into the input language.

11Returns a transcription object in `json`, `diarized_json`, or `verbose_json`

12format, or a stream of transcript events.

14### Parameters

16- `body AudioTranscriptionNewParams`

18 - `File param.Field[Reader]`

20 The audio file object (not file name) to transcribe, in one of these formats: flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.

22 - `Model param.Field[AudioModel]`

24 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`.

26 - `string`

28 - `type AudioModel string`

30 - `const AudioModelWhisper1 AudioModel = "whisper-1"`

32 - `const AudioModelGPT4oTranscribe AudioModel = "gpt-4o-transcribe"`

34 - `const AudioModelGPT4oMiniTranscribe AudioModel = "gpt-4o-mini-transcribe"`

36 - `const AudioModelGPT4oMiniTranscribe2025_12_15 AudioModel = "gpt-4o-mini-transcribe-2025-12-15"`

38 - `const AudioModelGPT4oTranscribeDiarize AudioModel = "gpt-4o-transcribe-diarize"`

40 - `ChunkingStrategy param.Field[AudioTranscriptionNewParamsChunkingStrategyUnion]`

42 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.

44 - `Auto`

46 - `const AutoAuto Auto = "auto"`

48 - `AudioTranscriptionNewParamsChunkingStrategyVadConfig`

50 - `Type string`

52 Must be set to `server_vad` to enable manual chunking using server side VAD.

54 - `const AudioTranscriptionNewParamsChunkingStrategyVadConfigTypeServerVad AudioTranscriptionNewParamsChunkingStrategyVadConfigType = "server_vad"`

56 - `PrefixPaddingMs int64`

58 Amount of audio to include before the VAD detected speech (in

59 milliseconds).

61 - `SilenceDurationMs int64`

63 Duration of silence to detect speech stop (in milliseconds).

64 With shorter values the model will respond more quickly,

65 but may jump in on short pauses from the user.

67 - `Threshold float64`

69 Sensitivity threshold (0.0 to 1.0) for voice activity detection. A

70 higher threshold will require louder audio to activate the model, and

71 thus might perform better in noisy environments.

73 - `Include param.Field[[]TranscriptionInclude]`

75 Additional information to include in the transcription response.

76 `logprobs` will return the log probabilities of the tokens in the

77 response to understand the model's confidence in the transcription.

78 `logprobs` only works with response_format set to `json` and only with

79 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`.

81 - `const TranscriptionIncludeLogprobs TranscriptionInclude = "logprobs"`

83 - `KnownSpeakerNames param.Field[[]string]`

85 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.

87 - `KnownSpeakerReferences param.Field[[]string]`

89 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`.

91 - `Language param.Field[string]`

93 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.

95 - `Prompt param.Field[string]`

97 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`.

99 - `ResponseFormat param.Field[AudioResponseFormat]`

100

101 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.

102

103 - ``

104

105 - `Temperature param.Field[float64]`

106

107 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.

108

109 - `TimestampGranularities param.Field[[]string]`

110

111 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.

112 This option is not available for `gpt-4o-transcribe-diarize`.

113

114 - `const AudioTranscriptionNewParamsTimestampGranularityWord AudioTranscriptionNewParamsTimestampGranularity = "word"`

115

116 - `const AudioTranscriptionNewParamsTimestampGranularitySegment AudioTranscriptionNewParamsTimestampGranularity = "segment"`

117

118### Returns

119

120- `type AudioTranscriptionNewResponseUnion interface{…}`

121

122 Represents a transcription response returned by model, based on the provided input.

123

124 - `type Transcription struct{…}`

125

126 Represents a transcription response returned by model, based on the provided input.

127

128 - `Text string`

129

130 The transcribed text.

131

132 - `Logprobs []TranscriptionLogprob`

133

134 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.

135

136 - `Token string`

137

138 The token in the transcription.

139

140 - `Bytes []float64`

141

142 The bytes of the token.

143

144 - `Logprob float64`

145

146 The log probability of the token.

147

148 - `Usage TranscriptionUsageUnion`

149

150 Token usage statistics for the request.

151

152 - `type TranscriptionUsageTokens struct{…}`

153

154 Usage statistics for models billed by token usage.

155

156 - `InputTokens int64`

157

158 Number of input tokens billed for this request.

159

160 - `OutputTokens int64`

161

162 Number of output tokens generated.

163

164 - `TotalTokens int64`

165

166 Total number of tokens used (input + output).

167

168 - `Type Tokens`

169

170 The type of the usage object. Always `tokens` for this variant.

171

172 - `const TokensTokens Tokens = "tokens"`

173

174 - `InputTokenDetails TranscriptionUsageTokensInputTokenDetails`

175

176 Details about the input tokens billed for this request.

177

178 - `AudioTokens int64`

179

180 Number of audio tokens billed for this request.

181

182 - `TextTokens int64`

183

184 Number of text tokens billed for this request.

185

186 - `type TranscriptionUsageDuration struct{…}`

187

188 Usage statistics for models billed by audio input duration.

189

190 - `Seconds float64`

191

192 Duration of the input audio in seconds.

193

194 - `Type Duration`

195

196 The type of the usage object. Always `duration` for this variant.

197

198 - `const DurationDuration Duration = "duration"`

199

200 - `type TranscriptionVerbose struct{…}`

201

202 Represents a verbose json transcription response returned by model, based on the provided input.

203

204 - `Duration float64`

205

206 The duration of the input audio.

207

208 - `Language string`

209

210 The language of the input audio.

211

212 - `Text string`

213

214 The transcribed text.

215

216 - `Segments []TranscriptionSegment`

217

218 Segments of the transcribed text and their corresponding details.

219

220 - `ID int64`

221

222 Unique identifier of the segment.

223

224 - `AvgLogprob float64`

225

226 Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.

227

228 - `CompressionRatio float64`

229

230 Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.

231

232 - `End float64`

233

234 End time of the segment in seconds.

235

236 - `NoSpeechProb float64`

237

238 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.

239

240 - `Seek int64`

241

242 Seek offset of the segment.

243

244 - `Start float64`

245

246 Start time of the segment in seconds.

247

248 - `Temperature float64`

249

250 Temperature parameter used for generating the segment.

251

252 - `Text string`

253

254 Text content of the segment.

255

256 - `Tokens []int64`

257

258 Array of token IDs for the text content.

259

260 - `Usage TranscriptionVerboseUsage`

261

262 Usage statistics for models billed by audio input duration.

263

264 - `Seconds float64`

265

266 Duration of the input audio in seconds.

267

268 - `Type Duration`

269

270 The type of the usage object. Always `duration` for this variant.

271

272 - `const DurationDuration Duration = "duration"`

273

274 - `Words []TranscriptionWord`

275

276 Extracted words and their corresponding timestamps.

277

278 - `End float64`

279

280 End time of the word in seconds.

281

282 - `Start float64`

283

284 Start time of the word in seconds.

285

286 - `Word string`

287

288 The text content of the word.

289

290### Example

291

292```go

293package main

294

295import (

296 "bytes"

297 "context"

298 "fmt"

299 "io"

300

301 "github.com/openai/openai-go"

302 "github.com/openai/openai-go/option"

303)

304

305func main() {

306 client := openai.NewClient(

307 option.WithAPIKey("My API Key"),

308 )

309 transcription, err := client.Audio.Transcriptions.New(context.TODO(), openai.AudioTranscriptionNewParams{

310 File: io.Reader(bytes.NewBuffer([]byte("Example data"))),

311 Model: openai.AudioModelGPT4oTranscribe,

312 })

313 if err != nil {

314 panic(err.Error())

315 }

316 fmt.Printf("%+v\n", transcription)

317}

318```

319

320#### Response

321

322```json

323{

324 "text": "text",

325 "logprobs": [

326 {

327 "token": "token",

328 "bytes": [

329 0

330 ],

331 "logprob": 0

332 }

333 ],

334 "usage": {

335 "input_tokens": 0,

336 "output_tokens": 0,

337 "total_tokens": 0,

338 "type": "tokens",

339 "input_token_details": {

340 "audio_tokens": 0,

341 "text_tokens": 0

342 }

343 }

344}

345```

346

347## Domain Types

348

349### Transcription

350

351- `type Transcription struct{…}`

352

353 Represents a transcription response returned by model, based on the provided input.

354

355 - `Text string`

356

357 The transcribed text.

358

359 - `Logprobs []TranscriptionLogprob`

360

361 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.

362

363 - `Token string`

364

365 The token in the transcription.

366

367 - `Bytes []float64`

368

369 The bytes of the token.

370

371 - `Logprob float64`

372

373 The log probability of the token.

374

375 - `Usage TranscriptionUsageUnion`

376

377 Token usage statistics for the request.

378

379 - `type TranscriptionUsageTokens struct{…}`

380

381 Usage statistics for models billed by token usage.

382

383 - `InputTokens int64`

384

385 Number of input tokens billed for this request.

386

387 - `OutputTokens int64`

388

389 Number of output tokens generated.

390

391 - `TotalTokens int64`

392

393 Total number of tokens used (input + output).

394

395 - `Type Tokens`

396

397 The type of the usage object. Always `tokens` for this variant.

398

399 - `const TokensTokens Tokens = "tokens"`

400

401 - `InputTokenDetails TranscriptionUsageTokensInputTokenDetails`

402

403 Details about the input tokens billed for this request.

404

405 - `AudioTokens int64`

406

407 Number of audio tokens billed for this request.

408

409 - `TextTokens int64`

410

411 Number of text tokens billed for this request.

412

413 - `type TranscriptionUsageDuration struct{…}`

414

415 Usage statistics for models billed by audio input duration.

416

417 - `Seconds float64`

418

419 Duration of the input audio in seconds.

420

421 - `Type Duration`

422

423 The type of the usage object. Always `duration` for this variant.

424

425 - `const DurationDuration Duration = "duration"`

426

427### Transcription Diarized

428

429- `type TranscriptionDiarized struct{…}`

430

431 Represents a diarized transcription response returned by the model, including the combined transcript and speaker-segment annotations.

432

433 - `Duration float64`

434

435 Duration of the input audio in seconds.

436

437 - `Segments []TranscriptionDiarizedSegment`

438

439 Segments of the transcript annotated with timestamps and speaker labels.

440

441 - `ID string`

442

443 Unique identifier for the segment.

444

445 - `End float64`

446

447 End timestamp of the segment in seconds.

448

449 - `Speaker string`

450

451 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`, ...).

452

453 - `Start float64`

454

455 Start timestamp of the segment in seconds.

456

457 - `Text string`

458

459 Transcript text for this segment.

460

461 - `Type TranscriptTextSegment`

462

463 The type of the segment. Always `transcript.text.segment`.

464

465 - `const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"`

466

467 - `Task Transcribe`

468

469 The type of task that was run. Always `transcribe`.

470

471 - `const TranscribeTranscribe Transcribe = "transcribe"`

472

473 - `Text string`

474

475 The concatenated transcript text for the entire audio input.

476

477 - `Usage TranscriptionDiarizedUsageUnion`

478

479 Token or duration usage statistics for the request.

480

481 - `TranscriptionDiarizedUsageTokens`

482

483 - `InputTokens int64`

484

485 Number of input tokens billed for this request.

486

487 - `OutputTokens int64`

488

489 Number of output tokens generated.

490

491 - `TotalTokens int64`

492

493 Total number of tokens used (input + output).

494

495 - `Type Tokens`

496

497 The type of the usage object. Always `tokens` for this variant.

498

499 - `const TokensTokens Tokens = "tokens"`

500

501 - `InputTokenDetails TranscriptionDiarizedUsageTokensInputTokenDetails`

502

503 Details about the input tokens billed for this request.

504

505 - `AudioTokens int64`

506

507 Number of audio tokens billed for this request.

508

509 - `TextTokens int64`

510

511 Number of text tokens billed for this request.

512

513 - `TranscriptionDiarizedUsageDuration`

514

515 - `Seconds float64`

516

517 Duration of the input audio in seconds.

518

519 - `Type Duration`

520

521 The type of the usage object. Always `duration` for this variant.

522

523 - `const DurationDuration Duration = "duration"`

524

525### Transcription Diarized Segment

526

527- `type TranscriptionDiarizedSegment struct{…}`

528

529 A segment of diarized transcript text with speaker metadata.

530

531 - `ID string`

532

533 Unique identifier for the segment.

534

535 - `End float64`

536

537 End timestamp of the segment in seconds.

538

539 - `Speaker string`

540

541 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`, ...).

542

543 - `Start float64`

544

545 Start timestamp of the segment in seconds.

546

547 - `Text string`

548

549 Transcript text for this segment.

550

551 - `Type TranscriptTextSegment`

552

553 The type of the segment. Always `transcript.text.segment`.

554

555 - `const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"`

556

557### Transcription Include

558

559- `type TranscriptionInclude string`

560

561 - `const TranscriptionIncludeLogprobs TranscriptionInclude = "logprobs"`

562

563### Transcription Segment

564

565- `type TranscriptionSegment struct{…}`

566

567 - `ID int64`

568

569 Unique identifier of the segment.

570

571 - `AvgLogprob float64`

572

573 Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.

574

575 - `CompressionRatio float64`

576

577 Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.

578

579 - `End float64`

580

581 End time of the segment in seconds.

582

583 - `NoSpeechProb float64`

584

585 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.

586

587 - `Seek int64`

588

589 Seek offset of the segment.

590

591 - `Start float64`

592

593 Start time of the segment in seconds.

594

595 - `Temperature float64`

596

597 Temperature parameter used for generating the segment.

598

599 - `Text string`

600

601 Text content of the segment.

602

603 - `Tokens []int64`

604

605 Array of token IDs for the text content.

606

607### Transcription Stream Event

608

609- `type TranscriptionStreamEventUnion interface{…}`

610

611 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`.

612

613 - `type TranscriptionTextSegmentEvent struct{…}`

614

615 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`.

616

617 - `ID string`

618

619 Unique identifier for the segment.

620

621 - `End float64`

622

623 End timestamp of the segment in seconds.

624

625 - `Speaker string`

626

627 Speaker label for this segment.

628

629 - `Start float64`

630

631 Start timestamp of the segment in seconds.

632

633 - `Text string`

634

635 Transcript text for this segment.

636

637 - `Type TranscriptTextSegment`

638

639 The type of the event. Always `transcript.text.segment`.

640

641 - `const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"`

642

643 - `type TranscriptionTextDeltaEvent struct{…}`

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 TranscriptTextDelta`

652

653 The type of the event. Always `transcript.text.delta`.

654

655 - `const TranscriptTextDeltaTranscriptTextDelta TranscriptTextDelta = "transcript.text.delta"`

656

657 - `Logprobs []TranscriptionTextDeltaEventLogprob`

658

659 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`.

660

661 - `Token string`

662

663 The token that was used to generate the log probability.

664

665 - `Bytes []int64`

666

667 The bytes that were used to generate the log probability.

668

669 - `Logprob float64`

670

671 The log probability of the token.

672

673 - `SegmentID string`

674

675 Identifier of the diarized segment that this delta belongs to. Only present when using `gpt-4o-transcribe-diarize`.

676

677 - `type TranscriptionTextDoneEvent struct{…}`

678

679 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`.

680

681 - `Text string`

682

683 The text that was transcribed.

684

685 - `Type TranscriptTextDone`

686

687 The type of the event. Always `transcript.text.done`.

688

689 - `const TranscriptTextDoneTranscriptTextDone TranscriptTextDone = "transcript.text.done"`

690

691 - `Logprobs []TranscriptionTextDoneEventLogprob`

692

693 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`.

694

695 - `Token string`

696

697 The token that was used to generate the log probability.

698

699 - `Bytes []int64`

700

701 The bytes that were used to generate the log probability.

702

703 - `Logprob float64`

704

705 The log probability of the token.

706

707 - `Usage TranscriptionTextDoneEventUsage`

708

709 Usage statistics for models billed by token usage.

710

711 - `InputTokens int64`

712

713 Number of input tokens billed for this request.

714

715 - `OutputTokens int64`

716

717 Number of output tokens generated.

718

719 - `TotalTokens int64`

720

721 Total number of tokens used (input + output).

722

723 - `Type Tokens`

724

725 The type of the usage object. Always `tokens` for this variant.

726

727 - `const TokensTokens Tokens = "tokens"`

728

729 - `InputTokenDetails TranscriptionTextDoneEventUsageInputTokenDetails`

730

731 Details about the input tokens billed for this request.

732

733 - `AudioTokens int64`

734

735 Number of audio tokens billed for this request.

736

737 - `TextTokens int64`

738

739 Number of text tokens billed for this request.

740

741### Transcription Text Delta Event

742

743- `type TranscriptionTextDeltaEvent struct{…}`

744

745 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`.

746

747 - `Delta string`

748

749 The text delta that was additionally transcribed.

750

751 - `Type TranscriptTextDelta`

752

753 The type of the event. Always `transcript.text.delta`.

754

755 - `const TranscriptTextDeltaTranscriptTextDelta TranscriptTextDelta = "transcript.text.delta"`

756

757 - `Logprobs []TranscriptionTextDeltaEventLogprob`

758

759 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`.

760

761 - `Token string`

762

763 The token that was used to generate the log probability.

764

765 - `Bytes []int64`

766

767 The bytes that were used to generate the log probability.

768

769 - `Logprob float64`

770

771 The log probability of the token.

772

773 - `SegmentID string`

774

775 Identifier of the diarized segment that this delta belongs to. Only present when using `gpt-4o-transcribe-diarize`.

776

777### Transcription Text Done Event

778

779- `type TranscriptionTextDoneEvent struct{…}`

780

781 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`.

782

783 - `Text string`

784

785 The text that was transcribed.

786

787 - `Type TranscriptTextDone`

788

789 The type of the event. Always `transcript.text.done`.

790

791 - `const TranscriptTextDoneTranscriptTextDone TranscriptTextDone = "transcript.text.done"`

792

793 - `Logprobs []TranscriptionTextDoneEventLogprob`

794

795 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`.

796

797 - `Token string`

798

799 The token that was used to generate the log probability.

800

801 - `Bytes []int64`

802

803 The bytes that were used to generate the log probability.

804

805 - `Logprob float64`

806

807 The log probability of the token.

808

809 - `Usage TranscriptionTextDoneEventUsage`

810

811 Usage statistics for models billed by token usage.

812

813 - `InputTokens int64`

814

815 Number of input tokens billed for this request.

816

817 - `OutputTokens int64`

818

819 Number of output tokens generated.

820

821 - `TotalTokens int64`

822

823 Total number of tokens used (input + output).

824

825 - `Type Tokens`

826

827 The type of the usage object. Always `tokens` for this variant.

828

829 - `const TokensTokens Tokens = "tokens"`

830

831 - `InputTokenDetails TranscriptionTextDoneEventUsageInputTokenDetails`

832

833 Details about the input tokens billed for this request.

834

835 - `AudioTokens int64`

836

837 Number of audio tokens billed for this request.

838

839 - `TextTokens int64`

840

841 Number of text tokens billed for this request.

842

843### Transcription Text Segment Event

844

845- `type TranscriptionTextSegmentEvent struct{…}`

846

847 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`.

848

849 - `ID string`

850

851 Unique identifier for the segment.

852

853 - `End float64`

854

855 End timestamp of the segment in seconds.

856

857 - `Speaker string`

858

859 Speaker label for this segment.

860

861 - `Start float64`

862

863 Start timestamp of the segment in seconds.

864

865 - `Text string`

866

867 Transcript text for this segment.

868

869 - `Type TranscriptTextSegment`

870

871 The type of the event. Always `transcript.text.segment`.

872

873 - `const TranscriptTextSegmentTranscriptTextSegment TranscriptTextSegment = "transcript.text.segment"`

874

875### Transcription Verbose

876

877- `type TranscriptionVerbose struct{…}`

878

879 Represents a verbose json transcription response returned by model, based on the provided input.

880

881 - `Duration float64`

882

883 The duration of the input audio.

884

885 - `Language string`

886

887 The language of the input audio.

888

889 - `Text string`

890

891 The transcribed text.

892

893 - `Segments []TranscriptionSegment`

894

895 Segments of the transcribed text and their corresponding details.

896

897 - `ID int64`

898

899 Unique identifier of the segment.

900

901 - `AvgLogprob float64`

902

903 Average logprob of the segment. If the value is lower than -1, consider the logprobs failed.

904

905 - `CompressionRatio float64`

906

907 Compression ratio of the segment. If the value is greater than 2.4, consider the compression failed.

908

909 - `End float64`

910

911 End time of the segment in seconds.

912

913 - `NoSpeechProb float64`

914

915 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.

916

917 - `Seek int64`

918

919 Seek offset of the segment.

920

921 - `Start float64`

922

923 Start time of the segment in seconds.

924

925 - `Temperature float64`

926

927 Temperature parameter used for generating the segment.

928

929 - `Text string`

930

931 Text content of the segment.

932

933 - `Tokens []int64`

934

935 Array of token IDs for the text content.

936

937 - `Usage TranscriptionVerboseUsage`

938

939 Usage statistics for models billed by audio input duration.

940

941 - `Seconds float64`

942

943 Duration of the input audio in seconds.

944

945 - `Type Duration`

946

947 The type of the usage object. Always `duration` for this variant.

948

949 - `const DurationDuration Duration = "duration"`

950

951 - `Words []TranscriptionWord`

952

953 Extracted words and their corresponding timestamps.

954

955 - `End float64`

956

957 End time of the word in seconds.

958

959 - `Start float64`

960

961 Start time of the word in seconds.

962

963 - `Word string`

964

965 The text content of the word.

966

967### Transcription Word

968

969- `type TranscriptionWord struct{…}`

970

971 - `End float64`

972

973 End time of the word in seconds.

974

975 - `Start float64`

976

977 Start time of the word in seconds.

978

979 - `Word string`

980

981 The text content of the word.