llmtranslate

Category: Natural Language Processing

License: MIT

llmtranslate is a lightweight Python library that leverages LangChain’s ChatModels (like OpenAI’s GPT‑4o and others) to perform language detection and full-text translation. It offers both synchronous and asynchronous interfaces, enabling developers to easily integrate multilingual translation functionality into Python applications or workflows. Installable via pip, it supports a wide range of languages and provides chunking and concurrency controls for handling large or multilingual texts.

Key Features

Detects input language with ISO‑639 codes and language name
Translates text to a specified target language
Supports multilingual inputs in a single document
Provides both synchronous (Translator) and asynchronous (AsyncTranslator) APIs
Controls chunk size (max_translation_chunk_length, max_translation_chunk_length_multilang)
Manages concurrency (max_concurrent_llm_calls, optional time-based limits)
Integrates seamlessly with LangChain ChatModels (OpenAI, Anthropic, Mistral, etc.)
Requires minimal setup: install LangChain model packages and llmtranslate
MIT‑licensed with stable releases (e.g., 0.7.20)

GitHub

Similar Projects

llmtranslate

Key Features

Similar Projects

Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context

AI Minimalist Translation Assistant (Chrome Extension)

LXMERT: Learning Cross-Modality Encoder Representations from Transformers

GPT-3: Generative Pretrained Transformer 3

ask‑multiple‑pdfs (MultiPDF Chat App)

Entmax: Sparse Alternative to Softmax for Neural Networks