Qwen 3 surpassed Claude 4 Opus and DeepSeek V3 in tests

The Qwen team released an update to their flagship model Qwen 3. The results are excellent. The new version surpasses such powerful models as Claude 4 Opus, Kimi K2 and DeepSeek V3 in many key indicators.

Particularly important is the developers’ strategic decision to abandon the hybrid approach. Now instead of combining Instruct and Reasoning modes in one model, they release separate specialized versions. Today the Instruct model is presented. And the version for deep reasoning is already in active development.

The foundation of the new architecture is Mixture of Experts technology. Where from 235 billion parameters only 22 billion are actively used. This makes the model significantly lighter for computations, which is critically important for real application.

The developers claim they significantly improved the model’s basic knowledge coverage, its logical reasoning capabilities and long context processing up to 256,000 tokens. The model now follows user preferences much better.

In the future, the team plans to distill technologies into younger versions, which will make Qwen 3’s power accessible not only to owners of top graphics cards.

AIvengo > Reviews > Qwen 3 surpassed Claude 4 Opus and DeepSeek V3 in tests

UBTech will send Walker S2 robots to serve on China's border for $37 million

Chinese company UBTech won a contract for $37 million. And will send humanoid robots Walker S2 to serve on China's border with Vietnam. South China Morning Post reports that the robots will interact with tourists and staff, perform logistics operations, inspect cargo and patrol the area. And characteristically — they can independently change their battery.

Anthropic accidentally revealed an internal document about Claude's "soul"

Anthropic accidentally revealed the "soul" of artificial intelligence to a user. And this is not a metaphor. This is a quite specific internal document.

Jensen Huang ordered Nvidia employees to use AI everywhere

Jensen Huang announced total mobilization under the banner of artificial intelligence inside Nvidia. And this is no longer a recommendation. This is a requirement.

AI chatbots generate content that exacerbates eating disorders

A joint study by Stanford University and the Center for Democracy and Technology showed a disturbing picture. Chatbots with artificial intelligence pose a serious risk to people with eating disorders. Scientists warn that neural networks hand out harmful advice about diets. They suggest ways to hide the disorder and generate "inspiring weight loss content" that worsens the problem.

OpenAGI released the Lux model that overtakes Google and OpenAI

Startup OpenAGI released the Lux model for computer control and claims this is a breakthrough. According to benchmarks, the model overtakes analogues from Google, OpenAI and Anthropic by a whole generation. Moreover, it works faster. About 1 second per step instead of 3 seconds for competitors. And 10 times cheaper in cost per processing 1 token.