LT Pipelines
Text Translation
Translate text files into any target language using an LLM.
The lt-translation pipeline translates the content of text files into a specified target language using a large language model. The source language is detected automatically unless explicitly provided.
200 credits per file · ~2 minutes
Accepted Files
| Type | MIME types |
|---|---|
| Text | text/* (plain text, CSV, …) |
Up to 5 files per run.
Context Parameters
| Parameter | Required | Description |
|---|---|---|
target_language | Yes | The language to translate into. Accepts language names ("English", "German") or ISO 639-1 codes ("en", "de"). |
source_language | No | The source language. Auto-detected by the model if omitted. |
Runs without a target_language value will fail. Always include it in contextData.
Output
{
"pipelineSlug": "lt-translation",
"output": {
"translatedText": "This is the translated content of the file.",
"sourceLanguage": "German",
"targetLanguage": "English",
"model": "claude"
},
"rawText": "This is the translated content of the file."
}Output Fields
| Field | Type | Description |
|---|---|---|
translatedText | string | Full translated text |
sourceLanguage | string | Detected or provided source language name |
targetLanguage | string | Target language name as interpreted by the model |
model | string | Model provider used ("claude" or "openai") |
rawText | string | Alias of translatedText |
API Example
POST /v2/lt-pipelines/project/{projectId}/run{
"pipelineSlug": "lt-translation",
"fileIds": ["file_abc123", "file_def456"],
"contextData": {
"target_language": "English",
"source_language": "German"
}
}Response:
{
"runIds": ["ltrun_aaa111", "ltrun_bbb222"]
}One run is created per file. Retrieve each result individually:
GET /v2/lt-pipelines/runs/ltrun_aaa111How is this guide?
Audio Transcription
Transcribe audio and video files to text with segment-level timestamps using Whisper.
OCR — Images & PDFs
Extract text from images and PDF documents using pdftotext for digital PDFs and a vision model for scanned documents.
© 2026 Crowdee GmbH. All rights reserved.