Google's Gemini 3 tops the leaderboards

PLUS: DeepSeek's million-dollar training fix, a more human Grok, and an AI presentation tool

Author

Gen AI Art
January 05, 2026 • Estimated Reading Time: 9 minutes

In partnership with

Google is making a major play for the top spot with the public release of its next-generation foundational model, Gemini 3.

The new model is already setting new performance records on key benchmarks for reasoning and multi-modal tasks. Does this release, along with its new developer tools, signal a major shift in the AI race and put Google back in the lead?

Today in AI:

  • Google's Gemini 3 tops the leaderboards

  • DeepSeek's million-dollar training fix

  • Grok gets a more human-like update

Stop Drowning In AI Information Overload

Your inbox is flooded with newsletters. Your feed is chaos. Somewhere in that noise are the insights that could transform your work—but who has time to find them?

The Deep View solves this. We read everything, analyze what matters, and deliver only the intelligence you need. No duplicate stories, no filler content, no wasted time. Just the essential AI developments that impact your industry, explained clearly and concisely.

Replace hours of scattered reading with five focused minutes. While others scramble to keep up, you'll stay ahead of developments that matter. 600,000+ professionals at top companies have already made this switch.

What’s new? Google has officially launched Gemini 3, its powerful next-generation foundational model, making it publicly available in the Gemini app and through developer tools.

What matters?

  • Gemini 3 is already topping leaderboards in reasoning and multi-modal tasks, showcasing its advanced capabilities across a range of real-world benchmarks.

  • The new model demonstrates its versatility by processing and understanding information across text, code, images, and video.

  • Alongside the model, Google released a public preview of Antigravity, a new AI coding studio that lets developers manage AI agents to build applications collaboratively.

Why it matters?

The release of Gemini 3 gives developers and creators access to a substantial increase in AI horsepower for building more capable applications. This signals Google’s strategy to not only advance its core models but also build an integrated ecosystem of tools that accelerates development.

GUIDE

What’s new? Chinese AI lab DeepSeek has published research on a new technique called mHC (Manifold-Constrained Hyper-Connections). This method solves critical stability problems that cause large-scale AI training runs to fail, making them more reliable and efficient.

What matters?

  • The core problem mHC solves is that advanced Transformer architectures often become unstable during training. Signals in the model's parallel data streams can either explode or vanish, crashing training runs that cost millions.

  • DeepSeek's solution adds mathematical “guardrails” that control these data streams, leading to significant performance gains. A researcher's analysis highlights a 7.2% improvement in complex reasoning with only a 6.7% increase in training time.

  • The paper’s release suggests a new flagship model may be coming soon. DeepSeek's founder is known for continuing his pattern of sharing major research just before a product launch, and the timing aligns with expectations for a release before Chinese New Year.

Why it matters?

This architectural improvement directly lowers the immense financial risk and time investment required to train cutting-edge AI. By making large model training more stable, this work helps make the next generation of more powerful and capable AI systems a practical reality.

PRESENTED BY MORNINGBREW

All the news that matters to your career & life

Hyper-relevant news. Bite-sized stories. Written with personality. And games that’ll keep you coming back.

Morning Brew is the go-to newsletter for anyone who wants to stay on top of the world’s most pressing stories — in a quick, witty, and actually enjoyable way. If it impacts your career or life, you can bet it’s covered in the Brew — with a few puns sprinkled in to keep things interesting.

Join over 4 million people who read Morning Brew every day, and start your mornings with the news that matters most — minus the boring stuff.

What’s new? xAI has released Grok 4.1, an upgraded model focused on delivering more nuanced, emotionally aware, and factually reliable conversations.

What matters?

  • The model significantly reduces factual errors and hallucinations, making it more reliable for prompts that seek information.

  • Professionals like writers and marketers can benefit from richer creative collaboration and smoother conversational outputs.

  • Users can adjust the model's style with two distinct settings—fast or deep—while its personality adapts to match user intent.

Why it matters?

Grok 4.1's focus on emotional nuance and reliability pushes conversational AI beyond just answering queries. This move positions it as a compelling alternative for users who need a more context-aware and human-feeling creative partner.

Everything else in AI

Futurepedia published a guide showing how to use Claude to instantly turn raw CSV files into clean, interactive dashboards without writing any code.

Yaara provides an AI writing assistant capable of generating original long-form content, social media posts, and persuasive ad copy using a suite of over 90 AI tools.

LetsEnhance uses AI to upscale image resolution to print-ready quality, fix blurry photos, and generate unique 4K art from text prompts.

Opus automates the text-to-video production pipeline, programmatically generating cinematic scenes with lighting, camera moves, and characters directly from a script.

Essential AI Guides - Reading List:

Let us know!

Work with us

Reach 100k+ engaged Tech Professionals, Engineers, Managers and decision makers. Join brands like MorningBrew, HubSpot, Prezi, Nike, Ahref, Roku, 1440, Superhuman, and others in showcasing your product to our audience. Get in touch now →