kdeps-tools-transcribe_audio_offline
v1.0.1Transcribe an audio file to text offline with whisper.cpp (whisper-cli): no API key, no network. Supports flac, mp3, ogg, wav. Uses the GGML model at modelPath, or ggml-base.en.bin in the kdeps models dir. The model is not downloaded: place the GGML file there first. Returns the transcribed text.
set tools / shelf tools
LLM context: ~161 tokens when loaded (prompt or tool definition)
Where it runs
Part of the tools set, used in workflow and agent mode. What the model is told about each built-in tool: its description, usage notes and parameters. kdeps runs the tool itself.
- the agent loop builds the request
- the tool definitions are sent to the modeltools/transcribe_audio_offline
- the model calls a tool
- kdeps runs it
- the result goes back to the model
Loaded by: kdeps (agent loop), /update, /settings, kdeps run, kdeps install, kdeps list, kdeps uninstall, kdeps update
In the code: kdeps (agent loop)agent.RegisterBuiltinToolsagent.definedAs
Install
A kdeps asset: installed with the first-run profile that includes its set, or the first time a command needs it, and updated without a new release.
Pin a version with kdeps install tools/transcribe_audio_offline@<version>.
Contents
Transcribe an audio file to text offline with whisper.cpp (whisper-cli): no API key, no network. Supports flac, mp3, ogg, wav. Uses the GGML model at modelPath, or ggml-base.en.bin in the kdeps models dir. The model is not downloaded: place the GGML file there first. Returns the transcribed text.
| Parameter | Type | Description |
|---|---|---|
| file * | string | Absolute path to the audio file to transcribe |
| language | string | ISO-639-1 code. Default: endefault: en |
| modelPath | string | Absolute path to a GGML model file. Default: <models dir>/ggml-base.en.bin |
Command (declared in YAML, no Go code)
whisper-cli
Arguments: -m{{or .args.modelPath (printf "%s/ggml-base.en.bin" (or .env.MODELS (printf "%s/.kdeps/models" .env.HOME)))}}-f{{.args.file}}-l{{.args.language}}-nt-np
Paths kept in the workspace: filemodelPath
Timeout: 600s
Offered when set: KDEPS_MODELS_DIR | , HOME | USERPROFILE |
version: 1.0.1
description: 'Transcribe an audio file to text offline with whisper.cpp (whisper-cli): no API key, no network. Supports flac, mp3, ogg, wav. Uses the GGML model at modelPath, or ggml-base.en.bin in the kdeps models dir. The model is not downloaded: place the GGML file there first. Returns the transcribed text.'
parameters:
file:
type: string
description: Absolute path to the audio file to transcribe
required: true
language:
type: string
description: 'ISO-639-1 code. Default: en'
default: en
modelPath:
type: string
description: 'Absolute path to a GGML model file. Default: <models dir>/ggml-base.en.bin'
exec:
env:
MODELS: [KDEPS_MODELS_DIR, ""] # optional: overrides ~/.kdeps/models
HOME: [HOME, USERPROFILE, ""]
command: whisper-cli
install: whisper-cpp
paths: [file, modelPath]
args:
- "-m"
- '{{or .args.modelPath (printf "%s/ggml-base.en.bin" (or .env.MODELS (printf "%s/.kdeps/models" .env.HOME)))}}'
- "-f"
- "{{.args.file}}"
- "-l"
- "{{.args.language}}"
- "-nt"
- "-np"
timeout: 600s
Versions
| Version | Published | Status |
|---|---|---|
| 1.0.1 | 10/23/2026 | active |
| 1.0.0 | 10/13/2026 | active |
Details
- Author
- kdeps
- License
- Apache-2.0
- Latest Version
- 1.0.1
- Published
- 10/13/2026