DeepSeek R1 surpassed Qwen 3 and reduced gap with Gemini 2.5 Pro
Data on DeepSeek R1, which received a serious update, has arrived. And the results are impressive. The model now confidently surpasses its competitor Qwen 3 with 235 billion parameters. Although it still lags behind flagships like Gemini 2.5 Pro and O3, the gap has significantly narrowed. The main improvement is related to increased reasoning depth – now the model uses an average of 23,000 tokens to solve tasks, while the previous version was limited to 12,000. This ability for deeper analysis brought impressive results. For example, in the AIME test, accuracy grew from 70% to 87.5%. Besides impressive successes in benchmarks, the new version began hallucinating much less and significantly improved its capabilities in frontend development. Although it still has to grow to Claude’s level in this sphere.
I think within the next year we will see a new wave of large language model integration into knowledge distillation systems. Where giant models will act as “teachers” for compact versions. This will lead to rapid breakthrough in small model efficiency and their implementation in mobile devices.
Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.
Boston Dynamics released Orbit 5.0 — AI reduced inspections by 70%A cool update Orbit 5.0 for the Spot robot control platform from Boston Dynamics has been released. Which fundamentally changes the approach to industrial analytics and monitoring! The system now allows centralized control of entire robot fleets across multiple facilities, providing operators with detailed real-time analytics.
Abu Dhabi will spend $2.5 billion on AI-managed city by 2027Imagine a city where artificial intelligence takes care of every aspect of your life. This is not science fiction, but the near future of Abu Dhabi! Companies BOLD Technologies and My Aion are developing a unified platform Aion Sentia. Which will take control of all urban systems — from transport to healthcare and education.
4 Chinese engineers smuggled 80 TB of AI data in backpacks to MalaysiaAn incredible technological odyssey is unfolding right now! Chinese engineers found a surprisingly analog way to bypass digital restrictions. Imagine: 4 employees of a Chinese AI startup are flying from Beijing to Kuala Lumpur. And each carries 15 hard drives in their backpack! In total — 80 terabytes of data for neural network training.