Gemma 4 gets a stealth update that fixes tool calling bugs and truncated responses under the same name
Quick Answer
Google's Gemma 4 AI model received a stealth update enhancing performance on Nvidia Hopper GPUs, fixing tool calling bugs, and reducing truncated responses.
Quick Take
Google's Gemma 4 AI model received a stealth update enhancing performance on Nvidia Hopper GPUs, fixing bugs, and reducing truncated responses. The update boosts processing speed by 25-70% with Flash Attention 4 and improves agentic reasoning by up to 10.1% in telecommunications. Users can also increase OCR resolution with a configurable parameter.
Key Points
- Flash Attention 4 increases processing speed by 25-70%, reducing time to first token by 31%.
- Gemma 4 31B shows a 10.1% improvement in agentic reasoning for telecommunications.
- Users can manually adjust 'max_soft_tokens' from 280 to 1,120 for better OCR results.
- All parameter sizes in the model generation received updates, including the new 12B release.
- Community feedback criticized the update being labeled under the same 'Gemma 4' name.
📖 Reader Mode
~2 min read
Jul 16, 2026
Google shipped an update to its open AI model Gemma 4 that speeds up performance on Nvidia Hopper GPUs, fixes tool calling bugs, and addresses problems with truncated responses. Turning on Flash Attention 4 boosts the speed at which the model processes incoming prompts by 25 to 70 percent, according to Google. Time to first token drops by up to 31 percent. Google also fixed bugs in tool calling, the feature that lets the model trigger external tools on its own.
Google says it also cut down on cases where the model would cut answers short or return incomplete responses. For image processing, users can manually raise the "max_soft_tokens" parameter from 280 to 1,120 to get sharper OCR results and support resolutions up to 2.51 megapixels. Google put up an interactive configurator on Hugging Face for that. The published benchmarks only compare the 31B and E4B variants against their predecessors, but the Hugging Face repository shows that all parameter sizes in this model generation got updated, including the newest 12B release. The community has pushed back on Google shipping the update under the same "Gemma 4" name instead of tagging it as a separate version like "Gemma 4.1."
— Originally published at the-decoder.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.
