> ## Documentation Index > Fetch the complete documentation index at: https://aircaps.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Create a transcript > Start transcribing audio with A5S Async v1. The job runs asynchronously: poll [Get a transcript](/api-reference/get-transcript) or pass `webhook_url`. ## OpenAPI ````yaml POST /v1/transcripts openapi: 3.1.0 info: title: AirCaps AI API version: 1.0.0 description: >- Speech-to-text: A5S Async v1 (REST; English, Spanish, French, German, Portuguese, Hindi, Japanese, Korean and Chinese) and A5S Small V2 Streaming (WebSocket; English). Speaker diarization: AirCaps Diarization-v1 (REST). Speaker identification: AirCaps Speaker ID-v1 (REST, synchronous). servers: - url: https://api.aircaps.com security: - bearerAuth: [] tags: - name: Transcripts - name: Diarizations - name: Speakers - name: Files - name: Account paths: /v1/transcripts: post: tags: - Transcripts summary: Create a transcript description: >- Start transcribing audio with A5S Async v1. The job runs asynchronously: poll [Get a transcript](/api-reference/get-transcript) or pass `webhook_url`. operationId: createTranscript parameters: - name: Idempotency-Key in: header required: false description: >- Up to 255 characters. Makes retries safe: the same key and body return the original job. schema: type: string maxLength: 255 - name: wait in: query required: false description: >- Seconds to wait for the transcript to finish before answering (0 to 120). Answers as soon as it finishes, or with its current state when the time is up. schema: type: number minimum: 0 maximum: 120 default: 0 requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/CreateTranscriptRequest' example: model: a5s-async-v1 audio_url: https://example.com/meeting.mp3 language_code: en webhook_url: https://example.com/hooks/aircaps metadata: meeting_id: 42 responses: '201': description: >- Created. A retried request with the same `Idempotency-Key` returns the original transcript with `200`. content: application/json: schema: $ref: '#/components/schemas/Transcript' example: id: tr_01k6z3m4q8w7e5r2t1y0u9i8o7 object: transcript status: queued created_at: '2026-10-06T18:01:02.123Z' started_at: null completed_at: null audio_url: null file_id: file_01k6z3m4q8w7e5r2t1y0u9i8o7 language_code: en model: a5s-async-v1 audio_duration: null metadata: meeting_id: 42 webhook_url: https://example.com/hooks/aircaps webhook_status: null webhook_status_code: null error: null timing_mode: phrase headers: Location: description: URL of the new transcript. schema: type: string '400': description: >- Invalid request. Codes: `invalid_request`, `unsupported_language`, `unsupported_model`, `no_enrolled_speakers`, `speaker_not_enrolled`. content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: invalid_request_error code: unsupported_language message: >- language 'it' is not supported; supported: English ("en", "en-US", ...), "auto", and "es", "fr", "de", "pt", "hi", "ja", "ko", "zh" param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 '401': description: 'Missing or invalid API key. Codes: `unauthorized`.' content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: authentication_error code: unauthorized message: invalid API key param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 '403': description: >- The account may not do this. Codes: `account_suspended`, `account_disabled`, `usage_limit_reached`, `model_access_denied`. content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: permission_error code: model_access_denied message: >- This account does not have access to A5S Async v1. To get access, talk to sales at https://aircaps.ai/contact. param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 '404': description: 'Not found. Codes: `file_not_found`, `speaker_not_found`.' content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: invalid_request_error code: file_not_found message: file_id not found or expired param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 '409': description: >- Idempotency key reused with a different body. Codes: `idempotency_conflict`. content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: invalid_request_error code: idempotency_conflict message: Idempotency-Key was used with a different request param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 '429': description: 'Rate or queue limit reached. Codes: `rate_limited`, `queue_full`.' content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: rate_limit_error code: rate_limited message: too many requests; retry after the Retry-After interval param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 headers: Retry-After: description: Seconds to wait before retrying. schema: type: integer '500': description: >- Server error. Safe to retry (use the same Idempotency-Key). Codes: `internal_error`. content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: api_error code: internal_error message: internal error param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 '503': description: >- Temporarily unavailable. Retry after the Retry-After interval. Codes: `service_unavailable`. content: application/json: schema: $ref: '#/components/schemas/Error' example: error: type: api_error code: service_unavailable message: temporarily unavailable; please retry param: null request_id: req_01k6z3m4q8w7e5r2t1y0u9i8o7 headers: Retry-After: description: Seconds to wait before retrying. schema: type: integer components: schemas: CreateTranscriptRequest: type: object additionalProperties: false description: Provide exactly one of `audio_url` or `file_id`. properties: model: type: string enum: - a5s-async-v1 default: a5s-async-v1 description: The model. Any other value returns `400 unsupported_model`. audio_url: type: string maxLength: 8192 format: uri description: >- An http(s) URL to download the audio from (public or presigned), up to 5 GB, resolving to a public address. example: https://example.com/meeting.mp3 file_id: type: string maxLength: 64 description: >- An uploaded file, usable by any number of transcripts for 24 hours after upload. ID from [Create an upload URL](/api-reference/upload-url) or [Upload a file](/api-reference/upload-file). language_code: type: string maxLength: 16 default: en description: >- Language of the audio: `en` (or a regional English code such as `en-US`, `en-GB` or `en-IN`; for best results use the one that matches the speakers' locale or accent), `es`, `fr`, `de`, `pt`, `hi`, `ja`, `ko` or `zh`, or `auto` to detect it. A region on another language's code is accepted (`es-MX` is `es`). The transcript reports the code without its region. Any other value returns `400 unsupported_language`. example: en webhook_url: type: string maxLength: 2048 format: uri description: >- Called with a [webhook event](/api-reference/objects/webhook-event) when the job finishes. webhook_auth_header_name: type: string maxLength: 256 description: Name of a header to add to webhook requests. Requires `webhook_url`. example: Authorization webhook_auth_header_value: type: string maxLength: 4096 description: Value of that header. metadata: type: object description: >- Any JSON object up to 16 KB; echoed on the transcript and in webhooks. example: meeting_id: 42 speaker_identification: oneOf: - type: boolean - $ref: '#/components/schemas/SpeakerIdentificationOptions' description: >- Name the transcript's speakers from your [enrolled speakers](/guides/speaker-id) with AirCaps Speaker ID-v1 (enabled per account; `403 model_access_denied` otherwise). `true` uses the defaults; an object sets `threshold` and `speaker_ids`. See [Speaker names](/guides/async-transcription#speaker-names). example: threshold: 0.6 timing_mode: type: string enum: - phrase - region default: phrase description: Timestamp style for `words` and `utterances`. keyterms: type: array maxItems: 60 items: type: string maxLength: 50 description: >- Key terms: up to 60 words or short phrases that may be spoken in the audio, such as names, products and jargon (each at most 6 words and 50 characters), to help the model recognize them. The text stays lowercase with numbers spelled out, so a key term appears as spoken: "GPT-4" as "gpt four". See [Key terms](/guides/async-transcription#key-terms). example: - AirCaps - Dr. Patel - GPT-4 large_meeting: type: boolean default: false description: >- Set to `true` for recordings with many active speakers, typically more than 7 to 9, such as large team meetings, panels, town halls and hearings. Improves speaker attribution for these recordings and may take a little longer to process. Leave it off for typical meetings and conversations. See [Large meetings](/guides/async-transcription#large-meetings). Transcript: type: object required: - id - object - status - created_at - started_at - completed_at - audio_url - file_id - language_code - model - audio_duration - metadata - webhook_url - webhook_status - webhook_status_code - error properties: id: type: string description: Transcript ID. example: tr_01k6z3m4q8w7e5r2t1y0u9i8o7 object: type: string enum: - transcript example: transcript status: type: string enum: - queued - processing - completed - error description: >- `queued`: waiting for a processing slot. `processing`: being transcribed. `completed`: result fields present. `error`: see `error`. created_at: type: string format: date-time description: When the transcript was created (UTC). started_at: type: - string - 'null' format: date-time description: When processing started. completed_at: type: - string - 'null' format: date-time description: When the job finished (completed or error). audio_url: type: - string - 'null' description: The submitted `audio_url`, if any. file_id: type: - string - 'null' description: The submitted `file_id`, if any. language_code: type: string enum: - en - es - fr - de - pt - hi - ja - ko - zh - auto description: >- Language of the transcript: the code sent, without its region (`en-GB` is `en`), or `auto`. model: type: string enum: - a5s-async-v1 audio_duration: type: - number - 'null' description: Audio duration in seconds, once known. example: 3600.512 metadata: type: - object - 'null' description: Your `metadata`, echoed back. webhook_url: type: - string - 'null' description: Webhook target, if any. webhook_status: type: - string - 'null' enum: - pending - delivered - failed - null description: Webhook delivery state, if `webhook_url` was set. webhook_status_code: type: - integer - 'null' description: HTTP status of the last webhook delivery. error: oneOf: - $ref: '#/components/schemas/JobError' - type: 'null' description: Set when `status` is `error`. text: type: string description: >- Completed only. Full transcript: verbatim, lowercase, no punctuation. speakers: type: array items: $ref: '#/components/schemas/TranscriptSpeaker' description: >- Completed only. With `speaker_identification`, each speaker also has `speaker_id`, `name`, `score`, `matches` and `compared_duration`. utterances: type: array items: $ref: '#/components/schemas/Utterance' description: >- Completed only. Speaker turns, in time order; overlapping speech overlaps. words: type: array items: $ref: '#/components/schemas/Word' description: Completed only. Every word, in time order. timing_mode: type: string enum: - phrase - region description: The `timing_mode` used. keyterms: type: array items: type: string description: >- Only when the transcript was created with `keyterms`: the list used (duplicates removed). large_meeting: type: boolean description: 'Only when the transcript was created with `large_meeting: true`.' speaker_identification: $ref: '#/components/schemas/TranscriptSpeakerIdentification' description: >- Only when the transcript was created with `speaker_identification`: its settings and outcome. Error: type: object required: - error properties: error: type: object required: - type - code - message - request_id properties: type: type: string enum: - invalid_request_error - authentication_error - permission_error - rate_limit_error - api_error code: type: string description: Machine-readable code; see [Errors](/errors). message: type: string param: type: - string - 'null' description: The request field at fault, if any. request_id: type: - string - 'null' description: >- Also in the `x-request-id` header. Include it when contacting support. `null` on 502/504 from the edge. SpeakerIdentificationOptions: type: object additionalProperties: false properties: speaker_ids: type: array items: type: string maxItems: 1000 description: >- Compare with these enrolled speakers only, for example the people invited to the meeting. Each must exist and have an enrollment. Default: every enrolled speaker of the account. example: - spk_01k73a9x8w7v6u5t4s3r2q1p0n - spk_01k73a7m6n5b4v3c2x1z0l9k8j threshold: type: number minimum: 0 maximum: 1 default: 0.6 description: >- A speaker gets an enrolled speaker's name when that is their best match and the score is at least this. 0.5 names a few more speakers when everyone who may speak is enrolled. example: 0.6 JobError: type: object required: - code - message properties: code: type: string enum: - invalid_audio - audio_too_long - file_too_large - download_failed - file_not_found - usage_limit_reached - internal_error - canceled description: Why the job failed. Failed jobs never count toward usage. message: type: string description: Human-readable detail. TranscriptSpeaker: type: object required: - id - speaking_time properties: id: type: string example: speaker_0 speaking_time: type: number description: Seconds this speaker talks. example: 1402.11 speaker_id: type: - string - 'null' description: >- Only with `speaker_identification`: the enrolled speaker matched with this speaker (its best score reached `threshold`), or `null`. Two speakers can match the same enrolled speaker when one person was split in two. example: spk_01k73a9x8w7v6u5t4s3r2q1p0n name: type: - string - 'null' description: >- Only with `speaker_identification`: the matched enrolled speaker's name. example: Ana García score: type: - number - 'null' description: >- Only with `speaker_identification`: the matched enrolled speaker's score, or `null`. example: 0.781 matches: type: array items: $ref: '#/components/schemas/SpeakerMatch' description: >- Only with `speaker_identification`: the 3 best-scoring enrolled speakers, best first, whether or not they reached `threshold`. Empty when the speaker had too little speech to compare. compared_duration: type: number description: >- Only with `speaker_identification`: seconds of this speaker's audio that were compared. example: 60 Utterance: type: object required: - speaker - start - end - text - words properties: speaker: type: string description: >- Speaker label (`speaker_0`, `speaker_1`, … in order of first speech). example: speaker_0 speaker_name: type: - string - 'null' description: >- Only with `speaker_identification`: the name of the enrolled speaker this speaker was matched with, or `null`. example: Ana García start: type: number description: Start time in seconds. example: 0.48 end: type: number description: End time in seconds. example: 6.12 text: type: string description: Text of the utterance. example: hey good afternoon everybody words: type: array items: $ref: '#/components/schemas/Word' description: Words of the utterance. Word: type: object required: - text - start - end - speaker properties: text: type: string description: The word. example: hey start: type: number description: Start time in seconds. example: 0.48 end: type: number description: End time in seconds. example: 0.71 speaker: type: string description: Speaker label. example: speaker_0 TranscriptSpeakerIdentification: type: object required: - model - status - threshold - speaker_ids - speakers_compared - speakers_identified - error properties: model: type: string enum: - aircaps-speaker-id-v1 status: type: string enum: - pending - completed - failed - skipped description: >- `pending`: the transcript is not finished. `completed`: matched (possibly nobody). `failed`: speaker identification was unavailable; the transcript is complete without names (see `error`). `skipped`: the transcript failed. threshold: type: number example: 0.6 speaker_ids: type: - array - 'null' items: type: string description: The requested `speaker_ids`, or `null` for every enrolled speaker. speakers_compared: type: - integer - 'null' description: >- Enrolled speakers compared (those that existed and had enrollments when the transcript was processed). example: 12 speakers_identified: type: - integer - 'null' description: Transcript speakers that got a name. example: 2 error: type: - object - 'null' properties: code: type: string enum: - speaker_identification_failed message: type: string SpeakerMatch: type: object required: - speaker_id - name - score properties: speaker_id: type: string description: Enrolled speaker ID. example: spk_01k73a9x8w7v6u5t4s3r2q1p0n name: type: - string - 'null' description: The enrolled speaker's name. example: Ana García score: type: number description: >- Similarity of this speaker's voice to the enrolled speaker (-1 to 1). example: 0.781 securitySchemes: bearerAuth: type: http scheme: bearer description: 'Your API key: `Authorization: Bearer aircaps_sk_...`' ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.