Skip to main content

CLI tool to translate text with LLMs

Project description

trn: CLI tool for translating text using LLMs

Features

  • Multiple input sources: command line arguments, stdin, clipboard, URLs, or files
  • OpenAI, Anthropic, and Gemini models
  • Configurable via command line arguments or environment variables

Installation

With uv installed, you can run it without the installation: uvx trn will always run the latest version of the command. You can also install it the usual way:

uv tool install trn
# or
pip install trn

Getting started

First, set the key for your LLM provider. Example for Google Gemini:

uvx --with llm-gemini llm keys set gemini

Second, specify your target language via the TRN_TO_LANGUAGE environment variable or -t argument. The -t argument takes priority when both are set.

Third, provide text to translate in one of these ways:

  • command line arguments
  • standard input
  • clipboard

The tool checks for arguments first, then standard input, then the clipboard.

You can also translate web pages or local PDF/image files by providing a URL or file path.

Basic Usage

# Translate from clipboard to default language
trn

# Translate command line text into French
trn -t french Hello world

# Translate from stdin
echo "Hello world" | trn

# Translate a file
trn document.pdf

# Translate from URL
trn https://example.com/article

# Use custom LLM
trn -m gpt-4o-mini

Requirements

  • Python 3.12+
  • LLM API key configured
  • UV installed (optional, but highly recommended)

Configuration

Set environment variables for convenience:

export TRN_TO_LANGUAGE=spanish
# optionally:
export TRN_MODEL=gpt-4o-mini

Development

Run trn with local changes:

uv run trn ...

Run tests:

make test

All the options

> trn --help
> trn --help
usage: trn [-h] -t TO_LANGUAGE [-m MODEL] [-p PROMPT] [-a PROMPT_ADD] [-w WRAP]
           [-v] [-d]
           [text ...]

positional arguments:
  text                  Text to translate, or URL(s), or path to file(s)
                        (default: None)

options:
  -h, --help            show this help message and exit
  -t, --to-language TO_LANGUAGE
                        Target language for translation [env var:
                        TRN_TO_LANGUAGE] (default: None)
  -m, --model MODEL     LLM to use (run 'uvx llm models' for available models)
                        [env var: TRN_MODEL] (default: gemini-flash-latest)
  -p, --prompt PROMPT   Custom prompt for translation [env var: TRN_PROMPT]
                        (default: Translate the text (it can be in any language)
                        into {to_language}. Don't explain that the output is a
                        translation. Tell me if you don't recognize
                        '{to_language}' language. If there is a file attached,
                        translate the contents of the file. {prompt_add})
  -a, --prompt-add PROMPT_ADD
                        Text to append to the prompt [env var: TRN_PROMPT_ADD]
                        (default: )
  -w, --wrap WRAP       Wrap output at N chars (use 0 to disable wrapping) [env
                        var: TRN_WRAP] (default: 80)
  -v, --verbose         Enable verbose output [env var: TRN_VERBOSE] (default:
                        False)
  -d, --debug           Enable debug output [env var: TRN_DEBUG] (default:
                        False)

 In general, command-line values override environment variables which override
defaults.

usage: trn [-h] -t TO_LANGUAGE [-m MODEL] [-p PROMPT] [-a PROMPT_ADD] [-w WRAP] [-v] [-d] [text ...]

positional arguments: text Text to translate, or URL(s), or path to file(s) (default: None)

options: -h, --help show this help message and exit -t, --to-language TO_LANGUAGE Target language for translation [env var: TRN_TO_LANGUAGE] (default: None) -m, --model MODEL LLM to use (run 'uvx llm models' for available models) [env var: TRN_MODEL] (default: gemini-flash-latest) -p, --prompt PROMPT Custom prompt for translation [env var: TRN_PROMPT] (default: Translate the text (it can be in any language) into {to_language}. Don't explain that the output is a translation. Tell me if you don't recognize '{to_language}' language. If there is a file attached, translate the contents of the file. {prompt_add}) -a, --prompt-add PROMPT_ADD Text to append to the prompt [env var: TRN_PROMPT_ADD] (default: ) -w, --wrap WRAP Wrap output at N chars (use 0 to disable wrapping) [env var: TRN_WRAP] (default: 80) -v, --verbose Enable verbose output [env var: TRN_VERBOSE] (default: False) -d, --debug Enable debug output [env var: TRN_DEBUG] (default: False)

In general, command-line values override environment variables which override defaults.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

trn-0.1.11.tar.gz (6.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

trn-0.1.11-py3-none-any.whl (5.2 kB view details)

Uploaded Python 3

File details

Details for the file trn-0.1.11.tar.gz.

File metadata

  • Download URL: trn-0.1.11.tar.gz
  • Upload date:
  • Size: 6.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.27 {"installer":{"name":"uv","version":"0.9.27","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for trn-0.1.11.tar.gz
Algorithm Hash digest
SHA256 9d73384a016744f93954a6e8d682c964db64794b85e8630085f0bd6735cc204c
MD5 8cc023ac4e8c5ae8df3c6c2e620fc66b
BLAKE2b-256 6c1d2a1e65f2749b720727646eeebfb5d0b0f48488524bc6f02c890486d96ce9

See more details on using hashes here.

File details

Details for the file trn-0.1.11-py3-none-any.whl.

File metadata

  • Download URL: trn-0.1.11-py3-none-any.whl
  • Upload date:
  • Size: 5.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.27 {"installer":{"name":"uv","version":"0.9.27","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for trn-0.1.11-py3-none-any.whl
Algorithm Hash digest
SHA256 59b34553c871156bb9a40b64c9919e76a2459aea79c18065b955e8106a786886
MD5 85fcb59fc9874b35982b5dac4d662377
BLAKE2b-256 f46167367f8aaaf4190bb96ab302a76820aaa37bec8b6a52e29a91b00801fde4

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page