DeepSeek v4.1 Flash

🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, much efficient. 🔹 Introducing the smallest exemplary successful our caller architecture family, pinch autochthonal ocular understanding. 🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models. 1/6