Gemini 4 Hydrogen on High effort built this 3d world from an image in 5 minutes and for less than 1 USD, seems like a pretty impressive model, Google is cooking hard, it seems. available also here to view in CmdBench website.
using CmdBriefs lead management and seeing everybody working is just mesmerizing to watch, it's just so satisfying to watch how Opus 5.5 leads the whole crew to build milestones so quickly
Gemini 4 Hydrogen (High) by @GoogleDeepMind debuts #5 on Agent Arena with +14.8% net improvement across 6.5k real-world agentic sessions.
Hydrogen stands out for its speed and cost efficiency:
- Fastest model in the top 10: 5.3 min/task (median), 3x faster than #1 Claude Fable 5.2 and 2.4x faster than GPT-6 Astra
- Cost-efficient: $1.6 cost/task (median), 3.7x cheaper than GPT-6 Astra and 4.7x cheaper than Claude Fable 5.1. API token pricing: $1.5/$7.5 input/output per million tokens
Notably, it spends more output tokens per task: 41% more than Carbon, 2.8x more than Argon.
Strong across key signals:
- #2 Steerability: +21% (behind only Fable 5.2 at +33%; ahead of Carbon at +19% and Opus 5.5 at +18%)
- #4 Praise vs. Complaint: +32%
- #5 Confirmed Success: +15%
Ranking by task category:
- #3 in Chat: +19%
- #7 in Work: +13%
- #9 in Code: +12%
Hydrogen delivers competitive performance while completing tasks significantly faster and at a fraction of the cost of several leading models. Congrats to the @GoogleDeepMind team on another impressive release!
Googles latest Gemini 4 Hydrogen scores 54.3% between Fable 5.2 Low and Fable 5.2 Medium, with the price tag only 1.50$ input and 7.50$ for output M tokens.
Meet Gemini 4 Hydrogen.
It brings cutting-edge intelligence to everyone, with speed and efficiency for everyday coding and agentic work.
Across 4 thinking levels, it stands on its own or teams up with Carbon and Argon — 2.5X faster than 3.8 Flash, outpacing competing models in its class.
Hot take: the Mac mini was never the hard part.
The hard part is that every agent still lives in a terminal tab on the laptop you're typing on, and you're the one walking between them.
What's actually running on yours right now?
Jensen Huang has resigned as NVIDIA CEO, effective immediately.
33 years ago he co-founded the company. GPUs + CUDA became the base layer for almost every AI model we use today.
CFO Colette Kress is interim CEO. Wild end to one of tech's biggest founder runs.
Grok Bot v1.0 now released, been running a whole bot team for my apps: a chief of staff bot, product owners, SEO, ops, security
the part I care about most in 1.0 is bots taking initiative but still asking before sensitive actions
proactive is great. proactive without approvals is how you wake up to surprises
Introducing Grok Bot 1.0. AI teammates that work alongside your whole team.
Share Bots, collaborate with them in Slack and Teams, invite them to meetings, and watch them learn and become more helpful.