Artificial Analysis' Intelligence Index: Gemini 4 Hydrogen nearly ties Claude Sonnet 5.5 and GPT-6.1 Sol, at a third of Sonnet's cost but ~2.5x that of Sol (@artificialanlys)
Google's new Gemini 4 Hydrogen has Intelligence just below Claude Sonnet 5.5 and GPT-6.1 Sol, at a third of Sonnet's cost but ~2.5x that of GPT-6.1 Sol @GoogleDeepMind has released their third new model in six weeks. Gemini 4 Hydrogen introduces an xhigh reasoning level and scores 54 on the Artifi...
Google is rolling out its latest AI model, Gemini 4 Hydrogen, which replaces the previous 3.8 Flash model. The name change follows Google's trend of naming models after elements, though Hydrogen's purpose is more straightforward - it replaces the faster, more cost-effective Flash models most users in Google's ecosystem rely on. The rollout began with cybersecurity researchers, and while general availability was initially limited to those using Google's most expensive AI subscription, access has since been made more widely available.
Users can now access Hydrogen immediately, marking a significant improvement over the previous Flash model across Google's products. Google claims that Hydrogen represents a major upgrade, running in four thinking levels (low, medium, high, and the new xhigh), with xhigh being 2.5 times faster than Flash 3.8. Even in high mode, Hydrogen reduces fabricated claims by over 5 times compared to Flash 3.8, as per Google's analysis.
Users can also expect better code generation performance, with Hydrogen tripling the score of 3.8 Flash in Terminal-Bench. In some instances, Hydrogen even matches the performance of Google's Argon model, scoring 54 in the Artificial Analysis Intelligence Index, compared to Argon's 50.
Written by urgent.news from Ars Technica's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
- Google rolls out Gemini 4 Hydrogen, finally replacing 3.8 Flash arstechnica.com
