Gemini 4 Hydrogen

Arena.ai's weekly roundup shows Anthropic's Claude Fable 5.2 taking first place in Agent Arena with a 30.5% net improvement across 7.8k real-world sessions, while Google DeepMind's Gemini 4 Hydrogen becomes the fastest top-10 model at a typical 7.3 seconds per step. Meta's smallest model Muse Ember debuted in Code Arena WebDev, Thinky Machines' Inkling 2 Small leads US open models, and Baidu's ERNIE 5.5 Flash Preview nears second among open models with a 1M-token context.

Google released Gemini 4 Hydrogen, a fast agentic model that edges out Gemini 4 Argon on the Artificial Analysis Intelligence Index (51 vs. 50) at a lower price. Its key design is role-splitting: in Google Antigravity, an Orchestrator mode has Gemini 4 Carbon plan while Hydrogen edits files, updates packages and runs tests, handing control back on friction. It scores 57% on Terminal-Bench 4.1 and finishes tasks 2.5x faster, though at $1.50/$7.50 per 1M tokens it costs twice the 3.8 Flash intro rate.

NVIDIA officially announced that Jensen Huang has resigned as CEO, ending a 33-year tenure. The move follows New York Times and Forbes reports of multiple personal relationships, including one with a person accused of being a Chinese spy, and comes after reports that the board was preparing to replace him, making this the most consequential AI leadership shakeup since the Google DeepMind reshuffle.

Anthropic released Claude Sheets, giving agents more direct control over spreadsheets, while Google used Gemini 4 Hydrogen to paint and animate thousands of cells frame by frame in Google Sheets; Microsoft Office 365 remained down at the same time. The contrast highlights divergent spreadsheet strategies across the three vendors.

Google has launched Gemini 4 Hydrogen, a lightweight member of the Gemini 4 family succeeding Gemini 3.8 Flash. It delivers 2.5x faster task completion, raises Terminal-Bench 4.1 from 18.2% to 57%, cuts hallucinated statements by over 5x, and scores 51 on the Artificial Analysis Intelligence Index, edging past the flagship Gemini 4 Argon at 50. With a 1M-token context window priced at $1.5 per million input tokens, it can also pair with Gemini 4 Carbon in a large-model-orchestrates, small-model-executes agent setup.

Google released Gemini 4 Hydrogen, the cheapest member of the Gemini 4 family, at $1.50/M input and $7.50/M output tokens, only about 25% below its larger siblings. Carbon and Argon share identical pricing and are both scheduled to double on December 1, 2026, with Google distinguishing Argon by lower token spend per turn, leaving the pricing scheme notably confusing.

Artificial Analysis released its AA-Robotics v0.1 benchmark, where Gemini 4 Hydrogen is Google's top-ranked model at fourth overall, trailing Claude Fable 5.2, GPT-6 Astra and GPT-6.1 Sol. Hydrogen candidly reported its own failures, including a minor drink spill and an unclosed lid while tidying blocks.

Google has released Gemini 4 Hydrogen, a fast general-purpose model for everyday coding and agentic workflows that replaces Gemini 3.8 Flash and beats the larger Gemini 4 Argon on benchmarks. It offers four reasoning levels, a one-million-token context window for both input and output, and is priced at $1.50 per million input tokens and $7.50 per million output tokens, available now via the Gemini API, Gemini app, Google Search and Google Antigravity.

Google released Gemini 4 Hydrogen, calling it its biggest model improvement to Search yet, designed for fast execution across multi-step agentic workflows rather than deep reasoning. It completes tasks 2.5x faster than the prior fast model Gemini 3.8 Flash and even beats the stronger Gemini 4 Argon on one broad intelligence benchmark. The update is rolling out to all Search users globally and also lands in AI Studio, the Gemini API, the Gemini App and the Antigravity developer platform.

Google has launched Gemini 4 Hydrogen as the immediate successor to 3.8 Flash, opening access to all users and replacing the older model across its product lineup. It offers four thinking levels (low, medium, high, and the new xhigh), more than triples 3.8 Flash's Terminal-Bench score, and edges out the flagship Argon on the Artificial Analysis Intelligence Index with 54 versus 50.