> ## Documentation Index > Fetch the complete documentation index at: https://docs.aihubmax.com/llms.txt > Use this file to discover all available pages before exploring further. # OpenAI Format - Transcriptions > - Generic Transcriptions API reference for all OpenAI-compatible speech-to-text models - Convert audio files to text - Supported models: `whisper-1` (recommended), `gpt-4o-transcribe`, `gpt-4o-mini-transcribe` - Supports language hint, prompt for style guidance and multiple response formats - Model-specific fields (timestamp granularities, streaming, diarization, etc.) are documented in the "Model-Specific Parameters" section below ## Model-Specific Parameters The OpenAI transcription endpoint exposes different fields depending on the model. The request body above only documents fields **common to all models**. The following sections describe model-specific or model-restricted fields. ### Supported Models | Model ID | Description | | ----------------------------------- | ------------------------------------------------------------------------------------------------- | | `whisper-1` | Classic Whisper V2 model. Supports the broadest set of output formats and timestamp granularities | | `gpt-4o-transcribe` | High-accuracy transcription. Only `json` output. Streamable | | `gpt-4o-mini-transcribe` | Lightweight high-accuracy transcription. Only `json` output. Streamable | | `gpt-4o-mini-transcribe-2025-12-15` | Versioned snapshot of `gpt-4o-mini-transcribe` | | `gpt-4o-transcribe-diarize` | Transcription with speaker diarization. Use `diarized_json` to receive per-segment speaker labels | ### `response_format` Compatibility Matrix | Model | Supported formats | | ---------------------------------------------------------- | -------------------------------------------------------------------------------------- | | `whisper-1` | `json` / `text` / `srt` / `verbose_json` / `vtt` | | `gpt-4o-transcribe`, `gpt-4o-mini-transcribe`(-2025-12-15) | `json` only | | `gpt-4o-transcribe-diarize` | `json` / `text` / `diarized_json` (use `diarized_json` to receive speaker annotations) | ### `whisper-1`-Only Features * **`timestamp_granularities[]`** — array, allowed values: `word` / `segment`, default `[segment]` * Word / segment-level timestamp granularity * **Takes effect only when `response_format=verbose_json`** * Sent as repeated form field `timestamp_granularities[]` * gpt-4o-\* models cannot use this in practice (they only support `json`); `gpt-4o-transcribe-diarize` explicitly disallows it * **Streaming not supported**: `stream=true` is silently ignored on `whisper-1`. ### gpt-4o-\* Series Parameters Applies to `gpt-4o-transcribe`, `gpt-4o-mini-transcribe`, `gpt-4o-mini-transcribe-2025-12-15`. * **`include[]`** — array, allowed value: `logprobs` * Returns the log probabilities of each token, useful for assessing model confidence * **Only effective when `response_format=json`** * Not available on `whisper-1` or `gpt-4o-transcribe-diarize` * **`stream`** — boolean, default `false` * Streams transcription results via SSE (Server-Sent Events) * Ignored on `whisper-1` * **`chunking_strategy`** — `"auto"` string or `server_vad` object * Controls how the audio is split into chunks. If unset, the audio is transcribed as a single block * When `"auto"`: the server normalizes loudness and then uses VAD to choose chunk boundaries * When a `server_vad` object (manual VAD tuning): | Field | Type | Default | Description | | --------------------- | ------- | ------- | ---------------------------------------------------------------------------------------------------- | | `type` | string | — | **Required**, must be `"server_vad"` | | `prefix_padding_ms` | integer | `300` | Audio (ms) included before VAD-detected speech | | `silence_duration_ms` | integer | `200` | Silence (ms) used to detect end of speech. Shorter values respond faster but may cut on short pauses | ### `gpt-4o-transcribe-diarize`-Only Parameters Applies only to `gpt-4o-transcribe-diarize` (speaker-diarization model). * **`chunking_strategy`** — **Required** for inputs longer than 30 seconds (recommended: `"auto"`) * **`known_speaker_names[]`** — array, max 4 * Identifier list for known speakers (e.g. `customer`, `agent`) * Maps 1-to-1 with `known_speaker_references[]` * **`known_speaker_references[]`** — array, max 4 * Reference audio for each speaker, in [data URL](https://developer.mozilla.org/en-US/docs/Web/HTTP/Basics_of_HTTP/Data_URLs) format * Each sample must be 2-10 seconds * Same audio formats as the `file` field ### Fields Not Supported by `gpt-4o-transcribe-diarize` The following fields are **not available** on `gpt-4o-transcribe-diarize`: | Field | Note | | --------------------------- | ----------------------------------------------------- | | `prompt` | Style/continuation prompt not supported | | `timestamp_granularities[]` | Word / segment timestamp granularity not configurable | | `include[]` | Additional returns like `logprobs` not supported | | `stream` | Streaming output not supported | ## OpenAPI ````yaml openapi/en/openai-format-transcriptions.json POST /v1/audio/transcriptions openapi: 3.1.0 info: title: OpenAI Format - Transcriptions version: '1.0' servers: - url: https://api.aihubmax.com security: - BearerAuth: [] paths: /v1/audio/transcriptions: post: tags: - Text Series > OpenAI Format summary: OpenAI Format - Transcriptions description: > - Generic Transcriptions API reference for all OpenAI-compatible speech-to-text models - Convert audio files to text - Supported models: `whisper-1` (recommended), `gpt-4o-transcribe`, `gpt-4o-mini-transcribe` - Supports language hint, prompt for style guidance and multiple response formats - Model-specific fields (timestamp granularities, streaming, diarization, etc.) are documented in the "Model-Specific Parameters" section below operationId: openai-format-transcriptions requestBody: required: true content: multipart/form-data: schema: $ref: '#/components/schemas/TranscriptionRequest' responses: '200': $ref: '#/components/responses/TranscriptionSuccess' '400': $ref: '#/components/responses/BadRequest' '401': $ref: '#/components/responses/Unauthorized' '402': $ref: '#/components/responses/PaymentRequired' '403': $ref: '#/components/responses/Forbidden' '429': $ref: '#/components/responses/TooManyRequests' '500': $ref: '#/components/responses/InternalServerError' components: schemas: TranscriptionRequest: type: object required: - file - model properties: file: type: string format: binary description: > Audio file to transcribe **Notes:** - Uploaded via multipart/form-data - Supported formats: flac / mp3 / mp4 / mpeg / mpga / m4a / ogg / wav / webm model: type: string description: >- Speech-to-text model ID. Allowed values: `whisper-1`, `gpt-4o-transcribe`, `gpt-4o-mini-transcribe` example: whisper-1 language: type: string description: >- ISO-639-1 language code of the input audio (e.g. `en`, `zh`, `ja`). Supplying this improves accuracy and latency. example: en prompt: type: string description: >- Optional text to guide the model's style or to continue from a previous audio segment. The prompt should match the audio language. response_format: type: string description: Format of the transcription output enum: - json - text - srt - verbose_json - vtt default: json temperature: type: number description: >- Sampling temperature between 0 and 1. Higher values produce more random output; 0 lets the model auto-tune. minimum: 0 maximum: 1 default: 0 TranscriptionResponse: type: object required: - text properties: text: type: string description: Transcribed text example: The weather is nice today, let's go for a walk in the park. TranscriptionVerboseResponse: type: object required: - task - language - duration - text properties: task: type: string description: Task type enum: - transcribe language: type: string description: Detected language of the audio example: english duration: type: number description: Audio duration in seconds example: 5.32 text: type: string description: Full transcription text example: The weather is nice today, let's go for a walk in the park. words: type: array description: >- Word-level timestamps (returned only with whisper-1 + verbose_json; see "Model-Specific Parameters" below) items: type: object properties: word: type: string description: Word content example: weather start: type: number description: Start time in seconds example: 0.42 end: type: number description: End time in seconds example: 0.86 segments: type: array description: >- Segment-level details (returned only with whisper-1 + verbose_json; see "Model-Specific Parameters" below) items: type: object properties: id: type: integer description: Segment index example: 0 seek: type: integer description: Segment seek offset example: 0 start: type: number description: Segment start time in seconds example: 0 end: type: number description: Segment end time in seconds example: 5.32 text: type: string description: Segment text example: The weather is nice today, let's go for a walk in the park. tokens: type: array description: Tokens of the segment items: type: integer temperature: type: number description: Sampling temperature example: 0 avg_logprob: type: number description: Average log probability example: -0.32 compression_ratio: type: number description: Compression ratio example: 1.21 no_speech_prob: type: number description: No-speech probability example: 0.02 ErrorResponse400: type: object properties: error: type: object properties: message: type: string example: Invalid request parameters type: type: string example: invalid_request_error ErrorResponse401: type: object properties: error: type: object properties: message: type: string example: Invalid API key type: type: string example: authentication_error ErrorResponse403: type: object properties: error: type: object properties: message: type: string example: Access denied type: type: string example: permission_error ErrorResponse429: type: object properties: error: type: object properties: message: type: string example: Rate limit exceeded type: type: string example: rate_limit_error ErrorResponse500: type: object properties: error: type: object properties: message: type: string example: Internal server error type: type: string example: server_error responses: TranscriptionSuccess: description: Transcription response content: application/json: schema: oneOf: - $ref: '#/components/schemas/TranscriptionResponse' - $ref: '#/components/schemas/TranscriptionVerboseResponse' examples: default-json: summary: Default JSON response description: Returned when `response_format=json` (default) value: text: The weather is nice today, let's go for a walk in the park. verbose-json: summary: Verbose JSON response description: >- Returned by whisper-1 when `response_format=verbose_json` with word-level timestamp granularity enabled (see "Model-Specific Parameters" below) value: task: transcribe language: english duration: 5.32 text: The weather is nice today, let's go for a walk in the park. words: - word: weather start: 0.42 end: 0.86 - word: nice start: 1.12 end: 1.38 segments: - id: 0 seek: 0 start: 0 end: 5.32 text: >- The weather is nice today, let's go for a walk in the park. tokens: - 50364 - 1029 - 31935 temperature: 0 avg_logprob: -0.32 compression_ratio: 1.21 no_speech_prob: 0.02 text/plain: schema: type: string examples: text: summary: Plain text response description: Returned when `response_format=text` value: The weather is nice today, let's go for a walk in the park. srt: summary: SRT subtitle response description: Returned when `response_format=srt` value: | 1 00:00:00,000 --> 00:00:05,320 The weather is nice today, let's go for a walk in the park. vtt: summary: VTT subtitle response description: Returned when `response_format=vtt` value: | WEBVTT 00:00:00.000 --> 00:00:05.320 The weather is nice today, let's go for a walk in the park. BadRequest: description: Bad Request content: application/json: schema: $ref: '#/components/schemas/ErrorResponse400' Unauthorized: description: Unauthorized content: application/json: schema: $ref: '#/components/schemas/ErrorResponse401' PaymentRequired: description: Insufficient balance content: application/json: schema: $ref: '#/components/schemas/ErrorResponse400' Forbidden: description: Forbidden content: application/json: schema: $ref: '#/components/schemas/ErrorResponse403' TooManyRequests: description: Too Many Requests content: application/json: schema: $ref: '#/components/schemas/ErrorResponse429' InternalServerError: description: Internal Server Error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse500' securitySchemes: BearerAuth: type: http scheme: bearer description: | ## All APIs require Bearer Token authentication ## Add to request header: `Authorization: Bearer YOUR_API_KEY` ````