DeepSeek
DeepSeek is a Chinese AI lab, backed by the High-Flyer hedge fund, that shook the industry with open reasoning models trained at a fraction of Western costs. Its R1 release in early 2025 wiped hundreds of billions off US tech stocks and made training efficiency a global priority. All our DeepSeek coverage is collected here.

10 Free AI Tools to Replace Paid Data Science Software
DataRobot costs $50,000 a year, Tableau costs $900 per user. But Ollama, AutoGluon, PandasAI, and seven other open tools now fully replace paid analytics and…

Anthropic Banned Being Cruel to Claude, So I Built a Profanity Filter for My Prompts
Anthropic's new rules forbid insulting its models. One developer built Angry Switcher — a Claude Code plugin that intercepts profanity in prompts and cleans them up…

StoreBench: LLM Agents Run an Online Store and Lose to a Script — 49% vs. 97%
The StoreBench benchmark put seven top LLMs in charge of a clothing store. The best, DeepSeek-V4-Pro, passes 49% of tasks; a simple heuristic passes 97%; humans…

Meta releases Muse Spark 1.2 and resets its AI strategy with an open-source manifesto

Alibaba Releases Qwen3.8-Max, China's Largest Model and a Rival to Fable 5 and GPT

Chinese AI Lab Researchers Are Taking Over X While OpenAI and Anthropic Go Quiet

Chinese Models Sweep OpenRouter's Top 5: MiMo-V2.5 Leads With 10.5 Trillion Tokens

OpenAI Publishes 'Abundant Intelligence' Strategy: AI to Become More Accessible
OpenAI outlines 'abundant intelligence': through a full-stack approach, it aims to make AI more powerful, more affordable and more widely useful, all at once and…

Moonshot Is Using 20,000 Nvidia Chips From Alibaba to Train Its Kimi Model
Chinese AI startup Moonshot has leased a cluster of about 20,000 Nvidia chips from Alibaba to develop Kimi, despite US export restrictions on high-performance…

Xiaohongshu's UltraEP Balances MoE Experts in 300 µs, Reaching 94.3% of Ideal

Anthropic's Dario Amodei: Not Against Open-Weight Models, but Wary of Chinese AI

DeepSeek Pauses 480 Billion Yuan Funding Round After Founder's Remarks Leak

US Leans Toward Targeted Bans on Chinese Open-Weight Models, Not a Blanket Prohibition

PEARL: A 4B Model Beats DeepSeek-V3.2 685B at Optimization Modeling
PEARL translates natural-language problems into solver code and fixes it in a loop. Its 4B-parameter version outperformed the 685B-parameter DeepSeek-V3.2.

Thinky Releases Open Inkling 975B-A41B Model Under Apache 2.0 License
Thinky has released Inkling, a 975B-parameter multimodal MoE model with open weights under Apache 2.0, plus a compact 276B version, Inkling-Small.









