TranscriptFetch System API

openapi: 3.2.0
info:
  title: Transcriptfetch System API
  version: '1.0'
  description: 'Operations tagged System across 2 of this provider''s published API definitions: transcriptfetch-api-v1-openapi.json, transcriptfetch-api-v2-openapi.json. Each path carries the servers of the definition it was published in.'
servers:
- url: https://transcriptfetch.com
  description: Production
security:
- bearerAuth: []
tags:
- name: System
  description: Service health and metadata
paths:
  /api/v1/me:
    get:
      tags:
      - System
      summary: Validate your key & check balance
      description: Validates the API key and returns the account's remaining credit balance. Free - never billed. Useful for programmatic balance checks and as a credential test in integrations.
      operationId: me
      externalDocs:
        description: 'Full API reference: fields, response shape, and error codes'
        url: https://transcriptfetch.com/docs/endpoints
      responses:
        '200':
          description: Success
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/SuccessEnvelope'
              examples:
                example:
                  value:
                    ok: true
                    request_id: req_…
                    data:
                      kind: me
                      user_id: user_…
                      credits: 250
                    usage:
                      credits_spent: 0
                      balance: 250
                      bytes: 0
        '401':
          $ref: '#/components/responses/Error'
    servers:
    - url: https://transcriptfetch.com
      description: Production
  /api/v1/health:
    get:
      tags:
      - System
      summary: Health check
      description: Public liveness probe for uptime monitoring. Returns 200 whenever the API is serving. No authentication required and no credits used.
      operationId: healthCheck
      security: []
      externalDocs:
        description: 'Full API reference: fields, response shape, and error codes'
        url: https://transcriptfetch.com/docs/endpoints
      responses:
        '200':
          description: Success
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Health'
              examples:
                example:
                  value:
                    status: ok
                    service: transcriptfetch-api
                    version: 1.0.0
                    time: '2026-06-16T00:00:00.000Z'
    servers:
    - url: https://transcriptfetch.com
      description: Production
  /api/v2/me:
    get:
      tags:
      - System
      summary: Validate your key & check balance
      description: Validates the API key and returns the account's remaining credit balance. Free - never billed. Useful for programmatic balance checks and as a credential test in integrations.
      operationId: me
      externalDocs:
        description: 'Full API reference: fields, response shape, and error codes'
        url: https://transcriptfetch.com/docs/endpoints
      responses:
        '200':
          description: Success
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/SuccessEnvelope'
              examples:
                example:
                  value:
                    ok: true
                    request_id: req_…
                    data:
                      kind: me
                      user_id: user_…
                      credits: 250
                    usage:
                      credits_spent: 0
                      balance: 250
                      bytes: 0
        '401':
          $ref: '#/components/responses/Error'
    servers:
    - url: https://transcriptfetch.com
      description: Production
  /api/v2/health:
    get:
      tags:
      - System
      summary: Health check
      description: Public liveness probe for uptime monitoring. Returns 200 whenever the API is serving. No authentication required and no credits used.
      operationId: healthCheck
      security: []
      externalDocs:
        description: 'Full API reference: fields, response shape, and error codes'
        url: https://transcriptfetch.com/docs/endpoints
      responses:
        '200':
          description: Success
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Health'
              examples:
                example:
                  value:
                    status: ok
                    service: transcriptfetch-api
                    version: 2.0.0
                    time: '2026-06-16T00:00:00.000Z'
    servers:
    - url: https://transcriptfetch.com
      description: Production
components:
  schemas:
    MeData:
      type: object
      required:
      - kind
      properties:
        kind:
          type: string
          enum:
          - me
        user_id:
          type: string
        credits:
          type:
          - integer
          - 'null'
          description: Remaining balance, or null for unlimited (admin) accounts.
    VideoListData:
      type: object
      required:
      - kind
      properties:
        kind:
          type: string
          enum:
          - video_list
        source:
          type: string
        videos:
          type: array
          items:
            $ref: '#/components/schemas/Video'
        next_cursor:
          type:
          - string
          - 'null'
          description: Pass back as `cursor` for the next page; null when exhausted.
    Health:
      type: object
      properties:
        status:
          type: string
          enum:
          - ok
        service:
          type: string
        version:
          type: string
        time:
          type: string
          format: date-time
    AiFallback:
      type: object
      description: Attached to transcript failures. Distinguishes 'this video has no captions' from 'we could not check', and says whether transcribing the audio would still work.
      required:
      - available
      - captions_unavailable
      - message
      properties:
        available:
          type: boolean
          description: True when retrying the same request with ai_fallback:true would actually start a transcription.
        captions_unavailable:
          type: boolean
          description: True only when we reached the video and confirmed no caption track exists. False means we could not check, NOT that captions exist.
        unavailable_reason:
          type: string
          enum:
          - unsupported_input
          - no_speech_to_transcribe
          - content_inaccessible
          - retry_captions_first
          - insufficient_credits
          description: 'Present when available is false: why AI transcription cannot be used.'
        message:
          type: string
        retry_with:
          type: object
          description: Merge into the original request body to trigger the fallback.
          properties:
            ai_fallback:
              type: boolean
              enum:
              - true
        cost_credits:
          type: integer
          description: Credits charged on successful delivery. Failures are free. AI transcription bills by audio length, so this is an estimate from the measured media length when cost_estimated is true, and otherwise the one-block minimum.
        cost_estimated:
          type: boolean
          description: True when cost_credits came from the media's real duration; false when it is only the minimum, because the length was not known at that point.
        balance:
          type:
          - integer
          - 'null'
          description: Remaining balance; null for admins.
    SuccessEnvelope:
      type: object
      required:
      - ok
      - request_id
      - data
      - usage
      properties:
        ok:
          type: boolean
          enum:
          - true
        request_id:
          type: string
        data:
          oneOf:
          - $ref: '#/components/schemas/TranscriptData'
          - $ref: '#/components/schemas/VideoListData'
          - $ref: '#/components/schemas/MeData'
          discriminator:
            propertyName: kind
        usage:
          $ref: '#/components/schemas/Usage'
    Video:
      type: object
      properties:
        videoId:
          type: string
        title:
          type: string
        thumbnailUrl:
          type: string
        duration:
          type:
          - number
          - 'null'
        channel:
          type:
          - string
          - 'null'
    TranscriptData:
      type: object
      required:
      - kind
      properties:
        kind:
          type: string
          enum:
          - transcript
        video_id:
          type: string
        title:
          type:
          - string
          - 'null'
        thumbnailUrl:
          type:
          - string
          - 'null'
          description: Poster image for the video. TikTok and Instagram serve signed, expiring URLs, so copy the image rather than hotlinking it.
        diarized:
          type: boolean
          description: True when speaker diarization produced labels (podcast transcriptions only); segments then carry `speaker` ids.
        text:
          type:
          - string
          - 'null'
        segments:
          type:
          - array
          - 'null'
          items:
            $ref: '#/components/schemas/Segment'
        podcast:
          type: object
          description: Present only when the input was a podcast link. Spotify and Apple do not host podcast audio; both read the publisher's RSS feed, so the link is resolved to that feed and the episode's own audio file. These fields say which show and episode were matched, so you can verify the resolution was correct.
          properties:
            show:
              type:
              - string
              - 'null'
            episode:
              type:
              - string
              - 'null'
            published_at:
              type:
              - string
              - 'null'
              format: date-time
            feed_url:
              type:
              - string
              - 'null'
              format: uri
            audio_url:
              type: string
              format: uri
            resolved_via:
              type: string
              enum:
              - spotify
              - apple
              - rss
              - direct
    ErrorEnvelope:
      type: object
      required:
      - ok
      - request_id
      - error
      properties:
        ok:
          type: boolean
          enum:
          - false
        request_id:
          type: string
        error:
          type: object
          required:
          - code
          - message
          properties:
            code:
              type: string
              enum:
              - unauthorized
              - invalid_request
              - insufficient_credits
              - rate_limited
              - idempotency_conflict
              - unsupported_platform
              - upstream_unavailable
              - internal_error
              - live_stream
              - was_live
              - no_audio_stream
              - audio_too_long
              - no_speech
              - captions_disabled
              - no_captions
              - age_restricted
              - members_only
              - private
              - unavailable
              - bot_gate
              - rate_limited
              - ip_blocked
              - region_blocked
              - timeout
              - upstream_error
              - proxy_unavailable
              - connection
              - parse_error
              - invalid_input
              - drm_protected
              - unknown
            message:
              type: string
            issues:
              type: array
              description: Field-level validation problems, when applicable.
              items:
                type: object
        reason:
          type: string
          description: Structured failure reason on transcript endpoints (no_captions, captions_disabled, age_restricted, rate_limited, …).
        ai_fallback:
          $ref: '#/components/schemas/AiFallback'
    Usage:
      type: object
      properties:
        credits_spent:
          type: integer
        balance:
          type:
          - integer
          - 'null'
          description: Remaining balance, or null for unlimited (admin) accounts.
        bytes:
          type: integer
    Segment:
      type: object
      properties:
        start:
          type: number
          description: Start time in seconds.
        duration:
          type: number
          description: Cue duration in seconds.
        text:
          type: string
        speaker:
          type: integer
          description: Speaker id (0, 1, …) on diarized podcast transcripts only; absent everywhere else.
    VideoListData_2:
      type: object
      required:
      - kind
      properties:
        kind:
          type: string
          enum:
          - video_list
        source:
          type: string
          description: 'Which listing produced the page: channel, playlist or search.'
        platform:
          type: string
          enum:
          - youtube
          - tiktok
          - instagram
          - spotify
          - apple
          - rss
          description: 'Where the rows came from: the `platform` field on search, the platform detected from the URL on channel and playlist.'
        videos:
          type: array
          items:
            $ref: '#/components/schemas/Video_2'
        next_cursor:
          type:
          - string
          - 'null'
          description: Pass back as `cursor` for the next page; null when exhausted.
    ErrorBlock:
      type: object
      required:
      - code
      - number
      - message
      - docs
      description: 'Every failure carries this block. Branch on code (a stable string) or number (a stable integer whose thousands digit is the family: 1 request, 2 account, 3 input, 4 content, 5 transient, 9 ours; 5xxx means retry with backoff). At most one of retry_with, details, issues is present. Failures are never charged.'
      properties:
        code:
          type: string
          enum:
          - invalid_request
          - invalid_cursor
          - unauthorized
          - idempotency_conflict
          - not_found
          - insufficient_credits
          - batch_too_large
          - rate_limited
          - unsupported_platform
          - endpoint_platform_mismatch
          - podcast_feed_not_found
          - audio_ineligible
          - invalid_input
          - audio_too_long
          - drm_protected
          - private
          - unavailable
          - members_only
          - age_restricted
          - region_blocked
          - live_stream
          - was_live
          - no_captions
          - captions_disabled
          - no_speech
          - no_audio_stream
          - upstream_error
          - timeout
          - connection
          - proxy_unavailable
          - parse_error
          - upstream_unavailable
          - internal_error
          - unknown
        number:
          type: integer
          enum:
          - 1001
          - 1002
          - 1101
          - 1201
          - 1301
          - 2001
          - 2002
          - 2101
          - 3001
          - 3002
          - 3003
          - 3004
          - 3005
          - 3006
          - 3007
          - 4001
          - 4002
          - 4003
          - 4004
          - 4005
          - 4101
          - 4102
          - 4103
          - 4104
          - 4105
          - 4106
          - 5001
          - 5002
          - 5003
          - 5004
          - 5005
          - 5101
          - 9001
          - 9002
        message:
          type: string
          description: Prose for humans. Never parse it.
        docs:
          type: string
          format: uri
          description: https://transcriptfetch.com/docs/errors/<code>
        retry_with:
          type: object
          description: 'Present when a different request would succeed: the fields to change, e.g. { "mode": "audio" } to transcribe a captionless video, or { "endpoint": "/api/v2/transcripts/video" } for a platform this endpoint does not list.'
          additionalProperties: true
        details:
          type: object
          description: 'Structured specifics for the few codes that document one: batch_too_large { max, sent }, podcast_feed_not_found { reason, candidates? }.'
          additionalProperties: true
        issues:
          type: array
          description: Field-level validation problems (invalid_request only).
          items:
            type: object
    Video_2:
      type: object
      description: 'One row of a listing. `url` is accepted as-is by the transcript and batch endpoints. The platform is not repeated per row: the page''s `platform` says it.'
      properties:
        videoId:
          type: string
        url:
          type: string
          format: uri
        title:
          type: string
        duration:
          type:
          - number
          - 'null'
          description: Seconds, when the source reports it.
        channel:
          type:
          - string
          - 'null'
        publishedAt:
          type:
          - string
          - 'null'
          format: date-time
          description: Upload time. Exact for TikTok, Instagram and podcasts; approximate on YouTube, whose listings only say "2 days ago", so it is exact to the day for recent videos and up to a year off for old ones.
        stats:
          type:
          - object
          - 'null'
          description: Engagement counts where the source exposes them.
          properties:
            plays:
              type:
              - integer
              - 'null'
    TranscriptData_2:
      type: object
      required:
      - kind
      properties:
        kind:
          type: string
          enum:
          - transcript
        video_id:
          type: string
        title:
          type:
          - string
          - 'null'
        source:
          type: string
          enum:
          - captions
          - audio
          description: 'Where the words came from: an existing caption track, or AI transcription of the audio (billed by length, see usage.credits_spent).'
        thumbnailUrl:
          type:
          - string
          - 'null'
          description: Poster image for the video. TikTok and Instagram serve signed, expiring URLs, so copy the image rather than hotlinking it.
        diarized:
          type: boolean
          description: True when speaker diarization produced labels (podcast transcriptions only); segments then carry `speaker` ids.
        text:
          type:
          - string
          - 'null'
        segments:
          type:
          - array
          - 'null'
          items:
            $ref: '#/components/schemas/Segment'
        podcast:
          type: object
          description: Present only when the input was a podcast link. Spotify and Apple do not host podcast audio; both read the publisher's RSS feed, so the link is resolved to that feed and the episode's own audio file. These fields say which show and episode were matched, so you can verify the resolution was correct.
          properties:
            show:
              type:
              - string
              - 'null'
            episode:
              type:
              - string
              - 'null'
            published_at:
              type:
              - string
              - 'null'
              format: date-time
            feed_url:
              type:
              - string
              - 'null'
              format: uri
            audio_url:
              type: string
              format: uri
            resolved_via:
              type: string
              enum:
              - spotify
              - apple
              - rss
              - direct
    ErrorEnvelope_2:
      type: object
      required:
      - ok
      - request_id
      - error
      properties:
        ok:
          type: boolean
          enum:
          - false
        request_id:
          type: string
        error:
          $ref: '#/components/schemas/ErrorBlock'
  responses:
    Error:
      description: Error
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/ErrorEnvelope'
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: 'Send your API key as `Authorization: Bearer <key>`.'
externalDocs:
  description: Documentation
  url: https://transcriptfetch.com/docs
x-refined-from:
- transcriptfetch-api-v1-openapi.json
- transcriptfetch-api-v2-openapi.json