Skip to content

Transcribe

Protocol: JSON 1.1 Header: X-Amz-Target: Transcribe.<Action> Endpoint prefix: transcribe

Floci emulates a small Amazon Transcribe control-plane stub for applications that create, inspect, list, and delete transcription jobs or custom vocabularies. Jobs transition to COMPLETED immediately and vocabularies transition to READY immediately.

No real audio transcription is performed. StartTranscriptionJob accepts the media URI and returns a synthetic TranscriptFileUri in the response so AWS SDK and CLI clients can exercise their Transcribe control-flow code locally.

Supported Operations

Operation Notes
StartTranscriptionJob Creates an in-memory job and returns it as COMPLETED
GetTranscriptionJob Returns a stored job by TranscriptionJobName
ListTranscriptionJobs Lists jobs with optional Status, JobNameContains, and MaxResults filters
DeleteTranscriptionJob Removes a stored job
CreateVocabulary Creates an in-memory vocabulary and returns it as READY
GetVocabulary Returns a stored vocabulary by VocabularyName
ListVocabularies Lists vocabularies with optional StateEquals, NameContains, and MaxResults filters
DeleteVocabulary Removes a stored vocabulary

LanguageCode defaults to en-US and MediaFormat defaults to mp4 when they are omitted from StartTranscriptionJob.

Configuration

Variable Default Description
FLOCI_SERVICES_TRANSCRIBE_ENABLED true Enable or disable the service

Examples

export AWS_ENDPOINT_URL=http://localhost:4566
export AWS_DEFAULT_REGION=us-east-1
export AWS_ACCESS_KEY_ID=test
export AWS_SECRET_ACCESS_KEY=test

aws transcribe start-transcription-job \
  --transcription-job-name floci-demo \
  --media MediaFileUri=s3://my-bucket/audio.mp3 \
  --language-code en-US \
  --media-format mp3

aws transcribe get-transcription-job \
  --transcription-job-name floci-demo

aws transcribe create-vocabulary \
  --vocabulary-name floci-vocab \
  --language-code en-US

aws transcribe list-vocabularies \
  --name-contains floci

import boto3

client = boto3.client(
    "transcribe",
    endpoint_url="http://localhost:4566",
    region_name="us-east-1",
)

client.start_transcription_job(
    TranscriptionJobName="floci-demo",
    Media={"MediaFileUri": "s3://my-bucket/audio.mp3"},
    LanguageCode="en-US",
    MediaFormat="mp3",
)

job = client.get_transcription_job(TranscriptionJobName="floci-demo")
print(job["TranscriptionJob"]["TranscriptionJobStatus"])  # COMPLETED

Out of Scope

  • Real speech-to-text processing or audio file parsing.
  • Writing transcript JSON objects to S3.
  • Persistent job or vocabulary storage across restarts.
  • Streaming, medical transcription, call analytics, and other Transcribe APIs beyond the operations listed above.