TranslateGemma
An experiment running TranslateGemma, Google's open-source translation model released in January 2026, locally via Ollama. Walks through installing TranslateGemma, implementing a Bash script that calls the Ollama API directly, and building a hotkey translation environment integrated with Alfred Workflow. This experience led to the development of general-purpose LLM tools beyond translation (myllm-cli and My LLM).
1. Overview
On January 15, 2026, Google announced "TranslateGemma," an open-source translation model. Built on Gemma 3, it supports translation across 55 languages. A notab…
2. Building the Translation Automation
The TranslateGemma model page on Ollama lists three parameter sizes: 4b, 12b, and 27b. All have a 128K context window and accept both text and image input. Thre…
3. Building the Alfred Workflow
Place the script at /usr/local/bin/translate so Alfred can invoke it. Once the path is in your $PATH, you can also run it directly from the terminal as translat…
4. Conclusion
By combining TranslateGemma, a Bash script, and an Alfred Workflow, it is possible to build an environment where selecting text and pressing a hotkey completes…