SP
BravenNow
Google’s new Gemini Pro model has record benchmark scores—again
| USA | technology | ✓ Verified - techcrunch.com

Google’s new Gemini Pro model has record benchmark scores—again

#Gemini Pro 3.1 #Google AI #Benchmark scores #Language models #AI competition #Technology advancement #Multi-step reasoning

📌 Key Takeaways

  • Google released Gemini Pro 3.1 with record benchmark scores
  • The model represents a significant improvement over its predecessor
  • Industry experts have validated its superior performance through independent benchmarks
  • The release intensifies competition in the rapidly evolving AI landscape

📖 Full Retelling

Google released its newest Gemini Pro model, version 3.1, on Thursday, February 18, 2026, achieving record benchmark scores in the competitive AI landscape where tech companies are racing to develop increasingly powerful language models capable of complex agentic work. The new model is currently available as a preview and will be generally released soon according to Google. Industry observers have noted that Gemini 3.1 Pro represents a significant advancement over its predecessor, Gemini 3, which was already considered a highly capable AI tool when it was released in November 2025. Google shared statistics from independent benchmarks, including one called Humanity's Last Exam, demonstrating the model's superior performance compared to previous versions. Brendan Foody, CEO of AI startup Mercor, praised the new model, noting that 'Gemini 3.1 Pro is now at the top of the APEX-Agents leaderboard' in a social media post, highlighting how the model's impressive results demonstrate the rapid improvement of AI systems in handling real knowledge work.

🏷️ Themes

AI advancement, Technology competition, Model performance

📚 Related People & Topics

Google AI

Google division dedicated to AI

Google AI is a subsidiary of Google DeepMind dedicated to artificial intelligence. It was announced at Google I/O 2017 by CEO Sundar Pichai. This division has expanded its reach with research facilities in various parts of the world such as Zurich, Paris, Israel, and Beijing.

View Profile → Wikipedia ↗

Language model

Statistical model of language

A language model is a computational model that predicts sequences in natural language. Language models are useful for a variety of tasks, including speech recognition, machine translation, natural language generation (generating more human-like text), optical character recognition, route optimizati...

View Profile → Wikipedia ↗
Competition in artificial intelligence

Competition in artificial intelligence

Rivalry between companies, nations, and researchers in developing AI technologies

Competition in artificial intelligence refers to the rivalry among companies, research institutions, and governments to develop and deploy the most capable artificial intelligence (AI) systems. The competition spans multiple domains, including large language models (LLMs), autonomous vehicles, robot...

View Profile → Wikipedia ↗

Entity Intersection Graph

Connections for Google AI:

🌐 Military applications of artificial intelligence 1 shared
🌐 Ethics of artificial intelligence 1 shared
🏢 Anthropic 1 shared
🌐 Pentagon 1 shared
🌐 Micrometre 1 shared
View full profile

Mentioned Entities

Google AI

Google division dedicated to AI

Language model

Statistical model of language

 Competition in artificial intelligence

Competition in artificial intelligence

Rivalry between companies, nations, and researchers in developing AI technologies

Deep Analysis

Why It Matters

Google's Gemini 3.1 Pro sets new benchmark records, indicating rapid progress in LLM capabilities and intensifying competition among AI leaders.

Context & Background

  • Gemini 3.1 Pro released as preview
  • Outperforms previous Gemini 3 on independent tests like Humanity’s Last Exam
  • Tops APEX-Agents leaderboard
  • Competes with new models from OpenAI and Anthropic
  • Highlights growing focus on agentic work and multi-step reasoning

What Happens Next

Google plans a full release soon, and the model will likely be integrated into its products. The competitive pressure may push other companies to accelerate their own model rollouts.

Frequently Asked Questions

What is Gemini 3.1 Pro?

The latest version of Google's Gemini LLM, offering higher accuracy and performance than its predecessor.

How does it compare to other models?

It tops independent benchmarks like APEX and Humanity’s Last Exam, outperforming previous Gemini releases and rival models.

When will it be widely available?

Google announced a preview release now and plans a general release soon, likely within the next few months.

}
Original Source
In Brief Posted: 4:55 PM PST · February 19, 2026 Lucas Ropek Google’s new Gemini Pro model has record benchmark scores—again On Thursday, Google released the newest version of Gemini Pro, its powerful LLM. The model, 3.1, is currently available as a preview and will be generally released soon, the company said. Google’s new model may be one of the most powerful LLMs yet. Onlookers have noted that Gemini 3.1 Pro appears to be a big step up from its predecessor, Gemini 3—which, upon its release in November, was already considered a highly capable AI tool. On Thursday, Google also shared statistics from independent benchmarks—such as one called Humanity’s Last Exam—that showed it performing significantly better than its previous version. Gemini 3.1 Pro was also praised by Brendan Foody, the CEO of AI startup Mercor, whose benchmarking system, APEX, is designed to measure how well new AI models perform real professional tasks. “Gemini 3.1 Pro is now at the top of the APEX-Agents leaderboard,” Foody said in a social media post , adding that the model’s impressive results show “how quickly agents are improving at real knowledge work.” The release comes as the AI model wars are heating up , and tech companies continue to release increasingly powerful LLMs designed for agentic work and multi-step reasoning. Other major names—including OpenAI and Anthropic—have recently released new models as well. Techcrunch event Save up to $300 or 30% to TechCrunch Founder Summit 1,000+ founders and investors come together at TechCrunch Founder Summit 2026 for a full day focused on growth, execution, and real-world scaling. Learn from founders and investors who have shaped the industry. Connect with peers navigating similar growth stages. Walk away with tactics you can apply immediately. Offer ends March 13. Save up to $300 or 30% to TechCrunch Founder Summit 1,000+ founders and investors come together at TechCrunch Founder Summit 2026 for a full day focused on growth, execution, and real...
Read full article at source

Source

techcrunch.com

More from USA

News from Other Countries

🇬🇧 United Kingdom

🇺🇦 Ukraine