Search AI Products and News

The rapid advancement of Artificial Intelligence (AI) models has led to concerns about the true performance of these models, despite continuous improvements by developers. To address this, the Vector Institute, founded by Geoffrey Hinton, has released a research study, "Assessing the State of the Art," which provides a comprehensive evaluation of 11 leading open-source and closed-source models through an interactive leaderboard. The evaluation covers mathematics, general knowledge, and coding.

DeepCoder-14B, a newly released open-source AI coding tool, demonstrates performance comparable to leading models like o1 and o3-mini, pushing the boundaries of what's possible in AI-assisted software development.

A groundbreaking achievement in quantum computing has been reached with the successful fine-tuning of a billion-parameter AI model on Origin Quantum's Wukong, marking a significant milestone in the development of quantum-enhanced AI.

Google recently announced the official pricing for its Gemini 2.5 Pro API. For input tokens up to 200,000, the base price is $1.25 per million input tokens, approximately 9.1 RMB; while the price for every million output tokens is $10, approximately 72.9 RMB. For longer texts exceeding 200,000 input tokens, the input price adjusts to $2.50 per million, approximately 18.2 RMB.

Ever felt frustrated by a seemingly clueless AI, wishing you could make it think before acting? Now, Anthropic has given AI a major brain upgrade! They've equipped their star model, Claude, with a groundbreaking 'think tool,' enabling it to handle complex tasks not by blindly rushing in, but by pausing to carefully consider its decisions, much like a human. This isn't just about slowing things down; it's about...

The official version of HunYuan T1, based on large-scale reinforcement learning, has undergone specialized optimization for science and engineering challenges such as mathematics, logical reasoning, science, and code, resulting in a significant improvement in reasoning capabilities. On common benchmarks, such as the enhanced MMLU-PRO dataset for evaluating large language models, HunYuan T1 achieved an excellent score of 87.2.

Kunlun Wanwei has open-sourced its Skywork R1V multimodal reasoning model, marking a significant advancement in Chinese AI capabilities. This powerful model represents a major leap forward in the field.