DeepSeek is a Chinese AI research laboratory that has developed highly capable open-source language models including DeepSeek-V3 and DeepSeek-R1, notable for their efficiency and performance.
China will allow Alibaba, ByteDance, and DeepSeek to purchase Nvidia's H200 chips for AI model training, but total approvals are capped at fewer than 200,000 units, less than half of what the companie…
A researcher found that ChatGPT and Claude blocked legitimate Lassa fever vaccine research queries while DeepSeek provided full assistance, highlighting inconsistencies in AI safety safeguards that ma…
OpenCode released data on real-world AI model usage, cache ratios, and costs, showing DeepSeek's v4-flash leading with 8.2T tokens, followed by DeepSeek v4-pro and Zhipu's glm-5.2. The data reveals sh…
A developer built a local AI sports agent using a DeepSeek R1 distilled Qwen 32B model on a LazyCat AI Pod, but found that local models lack practical value due to slow responses, outdated knowledge, …
A developer has added a scheduled task feature to the Neovim plugin chat.nvim, enabling AI assistants to proactively initiate conversations at future times. The design uses libuv timers with event-dri…
Zhipu AI quietly raised the default concurrency limit on its GLM API, allowing developers to run more parallel requests without prior approval. The move addresses developer complaints about severe rat…
A developer explains that multi-model AI applications require more than just access to multiple models; they need comprehensive visibility into reliability across different workflows. The post argues …
A developer details how OpenRouter serves as an abstraction layer for integrating multiple LLM providers into production AI applications. By simply changing the base URL and model name, developers can…
China reaffirmed its support for the United Nations as the primary platform for global AI governance and strongly advocated for open-source AI at the UN’s first Global Dialogue on AI Governance in Gen…
Chinese AI startup MiniMax is building a 2.7-trillion-parameter large language model, which it plans to open-source, according to The Information. The model, internally called M3 Pro, would be China's…
A CTO evaluated four Chinese LLM families—DeepSeek, Qwen, Kimi, and GLM—for production use at scale, finding that DeepSeek V4 Flash offers the best overall value at $0.25/M tokens, while Kimi K2.5 exc…
N8n founder Jan Oberhauser argued at the Raise Summit in Paris that companies should avoid model lock-in by using an orchestration layer that can switch between AI providers, as dependency on a single…
The US government is preparing to investigate and potentially restrict American companies' use of Chinese-made AI models, citing concerns that they advance Beijing's narratives and censor dissent. The…
DeepSeek is designing its own AI inference chip, targeting the phase where trained models generate responses, with manufacturing partner SMIC under US export controls. The chip leverages DeepSeek's Mi…
Researchers from Zhejiang University and Alibaba demonstrated a new denial-of-service attack on large language models by deliberately inducing "overthinking" through logically inconsistent prompts, ca…
A developer spent three weeks comparing AI API prices and found that the cheapest viable models are Apache 2.0 and MIT licensed models from China, not the well-known proprietary names. The analysis re…
China is shifting from competing on frontier AI performance to releasing affordable, open-weight AI models for mass adoption, mirroring its strategy in manufacturing. This approach allows global users…
Beijing is considering restricting foreign access to China's top AI models, a move that could disrupt global AI processing and crypto markets reliant on affordable Chinese AI. The potential curbs, dis…
A developer evaluated AI coding tools including web interfaces, custom IDEs, and VSCode extensions, finding Claude Code and Codex to be the most consistent and powerful options as of July 2026. The ar…
AI inference costs have plummeted for commodity models, with GPT-4-class output dropping from $20 to $0.40 per million tokens since 2022, while frontier model prices have surged, creating a split mark…