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.
Chinese open-weight AI models now occupy the top five spots on OpenRouter by weekly token usage, threatening to turn America's premium models into expensive niche products. Moonshot AI's Kimi K3, rele…
Moonshot AI, a Beijing-based startup founded by Yang Zhilin, unveiled Kimi K3 on July 16-17, 2026, as the world's largest open-weight AI model with 2.8 trillion parameters, outperforming OpenAI's GPT-…
OpenAI released the GPT-5.6 model family last week, which comes in three sizes each with roughly five or six reasoning-effort settings, according to Sebastian Raschka. The article explains how reasoni…
Mozilla's inaugural State of Open Source AI report finds that 79% of developers use open models but only 51% have deployed one to production, revealing a 28-point deployment gap. The report, based on …
A developer outlines a three-layer testing approach—syntax, schema, and semantic validity—for ensuring reliable structured outputs across multiple AI models including GPT, Claude, Gemini, DeepSeek, Qw…
Schemity, a desktop ERD tool built by a developer, positions itself as an offline alternative to ChartDB, the cloud-based schema visualizer. Schemity stores diagrams as local JSON files, supports Git-…
Schemity, a desktop ERD tool, now offers a bring-your-own-key (BYOK) AI chat that sends schema design requests directly from the user's machine to their chosen provider—OpenAI, Claude, Gemini, or Deep…
A ByteDance employee in Beijing burns through nearly a billion AI tokens monthly, yet ranks low in department consumption, as Chinese tech giants including ByteDance, Alibaba Group Holding, and Tencen…
OpenAI released GPT-5.6 Sol on July 9, its most capable coding and cybersecurity model, alongside two smaller siblings Terra and Luna, with pricing at $5/$30, $2.50/$15, and $1/$6 per million tokens r…
A fork of the DwarfStar inference engine, andreaborio/ds4, aims to run large Mixture-of-Experts models like Qwen3.6-35B-A3B on 16–64 GB Apple Silicon Macs by using adaptive SSD streaming and Metal res…
Moonshot AI's Kimi K3 model, with 2.8 trillion parameters and a one-million-token context window, has topped six of seven frontend domains on Arena.ai's benchmarks, surpassing models from Anthropic, O…
A Forrester report released last week found that technology sovereignty will grow slowly over the next five years, with the US and China scoring 79% and 82% respectively on Forrester's index while 12 …
A developer combined three AI models using Hermes Agent's Mixture of Agents (MoA) to build a Kanban board for a YouTube creator. The system sent the same prompt to DeepSeek and MiniMax as reference mo…
Businesses including DoorDash, Airbnb, and Siemens are experimenting with cheaper Chinese AI models from startups like Moonshot AI and DeepSeek as costs for U.S. rivals such as OpenAI, Google, and Ant…
China's Moonshot AI revealed its Kimi K3 model is nearly as powerful as US frontier models Anthropic's Fable 5 and OpenAI's GPT-5.6 Sol, while costing significantly less at $3 per million input tokens…
Beijing-based startup Moonshot's new Kimi K3 model has topped the charts on Arena's ranking of front-end coding capability, surprising the U.S. tech industry and signaling that open-source Chinese AI …
Moonshot AI launched Kimi K3, a 2.8-trillion-parameter open-weight model, on Thursday, ranking alongside Claude Fable 5 and GPT-5.6 Sol on the Artificial Analysis Intelligence Index with a score of 57…
A developer applied a 1913 inventory formula to determine the optimal time to restart LLM coding sessions, finding that context behaves like inventory. By instrumenting ~150 real sessions across 1,016…
Moonshot AI released Kimi K3 on July 16, a 2.8-trillion-parameter open-weight model that beats Claude Fable 5 and GPT-5.6 Sol on specific benchmarks, including a Writing Elo score of 2,840 versus Fabl…
Beijing-based startup Moonshot AI unveiled Kimi K3 on July 16, a 2.8 trillion parameter multimodal AI model that it claims can rival offerings from OpenAI and Anthropic, with API pricing at $3 per mil…