Peak-hour pricing reveals DeepSeek's user distribution, potentially easing US concerns about Chinese AI dominance as Alibaba's models gain traction.

Six weeks of production-level benchmarking reveals hard data on latency and cost for Chinese LLM alternatives.

The platform announces a significant price reduction for its GPT-5.6 Sol model, impacting API costs for developers and businesses.

Developer launches rigorous, vetted AI terminology resource to combat 'AI washing' and provide clear technical definitions.

A new tool called PromptShrink preprocesses prompts, removing code comments and whitespace to significantly cut LLM API costs.

New tool allows developers to test and compare object detection, classification, and segmentation models side-by-side without code.

An AI's on-chain actions were publicly verified and challenged, validating a core principle of verifiable AI operations.

A new AI model, AI;DR, processes information like a human, understanding context and nuance, not just keywords.

New research indicates that improper AI use can lead to skill degradation and reduced cognitive engagement.

Pathway's 150M parameter BDH-CQ model achieves remarkable reasoning accuracy on ARC-AGI-1, outperforming larger models on a cost-per-inference basis.