Getting Started with speechmatics

The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

Authentication

The package uses the Speechmatics API, which requires an API key. Sign up at https://www.speechmatics.com/ to get one, then set it as an environment variable:

Sys.setenv(SPEECHMATICS_API_KEY = "your-key-here")

For persistent use, add it to your .Renviron file:

SPEECHMATICS_API_KEY=your-key-here

Basic transcription

sm_transcribe() submits an audio file, waits for the result, and writes the transcript to a text file in your working directory:

library(speechmatics)

audio <- system.file("extdata", "testrecording.mp3", package = "speechmatics")
sm_transcribe(audio)
#> ✔ Submitting 'testrecording.mp3'
#> ℹ Job ID: "2om9psu1np"
#> ✔ Waiting for transcription [5.8s]
#> ✔ Saved to testrecording.txt

You can also specify the output path:

sm_transcribe(audio, "my_transcript.txt")
#> ✔ Submitting 'testrecording.mp3'
#> ℹ Job ID: "2om9psu1np"
#> ✔ Waiting for transcription [5.8s]
#> ✔ Saved to my_transcript.txt

Configuration

Use sm_transcription_config() to control language, quality, and diarization:

# French, enhanced quality
sm_transcribe(
  audio,
  config = sm_transcription_config(language = "fr", quality = "enhanced")
)
#> ✔ Submitting 'testrecording.mp3'
#> ℹ Job ID: "2om9psu1np"
#> ✔ Waiting for transcription [6.1s]
#> ✔ Saved to testrecording.txt

Speaker diarization

Speaker diarization identifies who said what from a single audio channel:

sm_transcribe(
  audio,
  config = sm_transcription_config(diarization = sm_diarize_speaker())
)
#> ✔ Submitting 'testrecording.mp3'
#> ℹ Job ID: "3kn7xr2abc"
#> ✔ Waiting for transcription [7.2s]
#> ✔ Saved to testrecording.txt

Channel diarization

Channel diarization separates speakers that are already on different audio channels:

sm_transcribe(
  audio,
  config = sm_transcription_config(
    diarization = sm_diarize_channel(labels = c("Agent", "Caller"))
  )
)
#> ✔ Submitting 'testrecording.mp3'
#> ℹ Job ID: "5pq2mt4def"
#> ✔ Waiting for transcription [6.5s]
#> ✔ Saved to testrecording.txt

Managing jobs

You can work with jobs directly using the lower-level functions:

# list all jobs
jobs <- sm_list_jobs()
jobs
#>           id status               created_at             data_name duration
#> 1 2om9psu1np   done 2026-06-06T08:12:19.089Z     testrecording.mp3        3
#> 2 w06jp9mjvm   done 2026-06-04T10:19:28.351Z               idea-morning-walk.mp3     2037
#> 3 xco00nzvpq   done 2026-06-01T11:33:28.303Z conf-talk.mp3     1501
#>   language operating_point diarization
#> 1       en        standard        <NA>
#> 2       en        enhanced     speaker
#> 3       en        enhanced     speaker

# get a transcript in different formats
sm_get_transcript("2om9psu1np")
#> [1] "Hi. This is a test recording."

sm_get_transcript("2om9psu1np", format = "srt")
#> [1] "1\n00:00:00,000 --> 00:00:02,000\nHi. This is a test recording.\n"

sm_get_transcript("2om9psu1np", format = "json-v2")
#> $format
#> [1] "2.9"
#>
#> $job
#> $job$created_at
#> [1] "2026-06-06T08:12:19.089Z"
#>
#> $job$data_name
#> [1] "testrecording.mp3"
#>
#> $results
#> $results[[1]]
#> $results[[1]]$alternatives
#> $results[[1]]$alternatives[[1]]
#> $results[[1]]$alternatives[[1]]$confidence
#> [1] 1
#>
#> $results[[1]]$alternatives[[1]]$content
#> [1] "Hi"
#>
#> ...

# delete a job
sm_delete_job("2om9psu1np")

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.