Google Releases TranslateGemma for Open Translation
The new 4B-parameter model is an instruction-tuned variant of Gemma, designed specifically for high-quality multilingual translation tasks.
Google has introduced TranslateGemma, a new family of open models specialized for high-quality multilingual translation. The initial release is a 4-billion-parameter model that has been instruction-tuned, making it adept at following commands for direct and nuanced text translation between various languages.
Built upon the solid foundation of Google's Gemma architecture, TranslateGemma is not a general-purpose chat model. Its "IT" designation signifies that it has been specifically fine-tuned on a large dataset of translation tasks. This specialization aims to provide higher accuracy and efficiency for these use cases compared to more generalized large language models.
The release of a capable, open translation model from a major AI lab is significant. It gives developers and researchers a powerful tool that can be run locally or self-hosted, providing a strong alternative to proprietary translation APIs. Its relatively modest 4B size makes it accessible for experimentation on a wider range of hardware.
The model is available now on Hugging Face under the Gemma license, which outlines terms for commercial use and distribution.
Sources
- Visit
google/translategemma-4b-it
Hugging Face
More from Google DeepMind
All Google DeepMind releases →
EmbeddingGemma 2 Expands to Any Modality
Google DeepMind's compact embedding model now spans text, vision, audio, and video in a single vector space.
Google DeepMind's Gemma 4 Goes Multimodal and MoE
The new open-weights family adds a mixture-of-experts design, encoder-free multimodal inputs, and an optional thinking mode.
Google DeepMind Releases TabFM for Tabular Data
A new foundation model brings zero-shot, in-context learning to classification and regression on structured tables.
More in Text / LLM
All Text / LLM →IBM's Granite Switch Treats Models Like Software
A new Apache-2.0 release reframes enterprise LLMs as components you can swap and update in place.
Mistral Large 4 Arrives as a 1T-Parameter MoE
The new flagship pairs a trillion total parameters with 52B active per token and adds vision alongside multilingual text.
Nanbeige 4.3 Arrives as a 33B Bilingual Model
A dense English–Chinese language model lands under Apache 2.0, aimed at teams that want permissive licensing at mid-size scale.
0 comments
No comments yet. Be the first to weigh in.