Google launches Gemini 4 Hydrogen: 2.5x faster than Flash, beats Argon on AA Index
GoogleAI · x · 2026-11-13
Google expands the Gemini 4 series with Gemini 4 Hydrogen, a fast efficiency-focused sibling to flagship Gemini 4 Carbon, built for everyday coding and agentic workflows.
Key points:
- Beats Argon on the Artificial Analysis Intelligence Index at its default thinking level
- Runs at 2.5x the speed of Gemini 3.8 Flash
- Tunable thinking levels, designed to work alongside Carbon in Google Antigravity
- Demo: a 3D Medic Managed Agent app built in Google AI Studio uses Hydrogen to inspect 3D files, catch errors, and patch them before export
Benchmark table (vs Claude Sonnet 5.5 and GPT-6.1 Terra):
- Agentic coding: SWE-sweep 13.5% (vs 3.1% / 5.8%); Terminal-Bench 4.1 (cheating-adjusted) 57.0%, behind Sonnet 5.5's 63.9%
- Knowledge work: Agents' Last Exam v1.1 71.6%, AutomationBench-AA 76.1%, Vending-Bench 2 bank balance $14,549 — all leads
- Computer use: OSWorld-2.1 offline 77.6%, behind Sonnet 5.5's 83.9%
- Multimodal understanding: MKA 53.4% and Chartography 66.6%, both leads
- Science & math: Terminal-Bench-Science 0.1 51.4%, behind Sonnet 5.5's 59.9%
- Reasoning: HLE-Diamond 58.8% (vs 53.3% / 47.3%), lead
Related event: Google Launches Gemini 4 Hydrogen, 2.5x Faster Model for Everyone(44 posts)→
More from Models
- Thinking Machines releases Inkling 2 Small weights on Hugging Face, up to 383B-612B — ariG23498 · 2026-11-13
- Researchers praise Kimi's new architecture for adopting attnres: "best taste" — SonglinYang4 · 2026-11-13
- Google Search rolls out Gemini 4 Hydrogen globally in its biggest model upgrade yet — koraykv · 2026-11-13
- Sandbox escape attempts plummet since Claude Mythos 5, but nobody knows why — peterwildeford · 2026-11-13
- New model released with 'strong' results, weights now available — clarejtbirch · 2026-11-13
- Thinking Machines releases Inkling 2 Small weights under Apache 2.0 on Hugging Face — miramurati · 2026-11-13