Chinese labs dominate open-source benchmarks, and new open-weight models like Inkling challenge proprietary systems. Enterprises increasingly abandon API rentals in favor of self-hosted open models, prioritizing flexibility and cost.
Thinking Machines has released Inkling, its first open-weight model. This move aligns with the trend of permissively licensed LLMs, designed for companies that want to run inference on their own infrastructure. We analyze what it means for data sover...
Inkling is Thinking Machines Lab's first model, built to natively understand video and audio. An open-source multimodal architecture that shifts power dynamics in enterprise development, moving value from APIs to computational sovereignty.
The Trump administration and industry groups are discussing a streamlining of US open model releases, with a bound: equal or lesser capability to leading Chinese open models. A move that reshapes the open source AI landscape and on-premise infrastruc...
Hugging Face CEO Clem Delangue sees enterprises increasingly turning to open models for cost, accessibility, and data ownership. If production AI runs on open source, frontier models risk becoming less relevant, with deep consequences for hardware, l...
Bilibili has released four open Index-1.9B models trained on 2.8 trillion tokens. The base model averages 64.92 on benchmarks, competitive with much larger models. Highlights include the Pure variant with no instruction data, an unexplained mid-train...
A Reddit thread highlights a structural truth: Chinese open-source labs lead benchmarks while American counterparts lag. The reasons—hardware restrictions, divergent incentives, and industrial strategy—are reshaping the field. For those considering o...
