Rocket - Start building, Now!
Mini toolsT5Gemma 2
By Google
Model family: Gemma
T5Gemma 2 is an open family of encoder-decoder models adapted from Gemma 3 using the UL2 objective. It accepts text and image inputs, generates text output, and is released in 270M-270M, 1B-1B and 4B-4B sizes with a 128K token context window across 140+ languages. Architectural changes like tied word embeddings and merged decoder self and cross attention cut parameters while keeping strong performance, giving compact models that often match or beat similar size Gemma 3 checkpoints on reasoning, code and long-context benchmarks.
New
Text
Gen 7
Overview
T5Gemma 2 is Google’s next generation Gemma 3 based encoder-decoder family, a lightweight multilingual and multimodal LLM that reads text and images, outputs text, and offers 128K context with tied embeddings and merged attention for efficient on-device deployment.
About Google
At Google, we think that AI can meaningfully improve people's lives and that the biggest impact will come when everyone can access it.
View Company ProfileTools using T5Gemma 2
No tools found for this model yet.
Last updated: February 25, 2026
