Google's TranslateGemma 4B It is a chat model designed for translation tasks across 55 languages, capable of handling text and image inputs with a total context length of 2K tokens. It is genuinely best at providing state-of-the-art translation capabilities in a lightweight and open-source package, making it accessible for deployment in resource-limited environments.
Input
Output
Context
2K
Max Output
-
Parameters
4B
Input Modalities
Output Modalities
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Data sourced from official provider APIs and documentation
Last updated: Jun 24, 2026
Automatically route workloads to the right model for every task, every time.