On April 10, the DeepSeek official blog published an article introducing DeepSeek V4, the flagship model set to be launched by DeepSeek. This model not only breaks the limits of parameter scale but also promises unprecedented efficiency. DeepSeek V4 is expected to handle 1 trillion (1T) parameters, natively supports multimodal data including text, images, videos, and audio, and features a context window of 1 million tokens (equivalent to 15-20 complete novels), making it a direct competitor to Western giants like OpenAI's GPT-5.4 and Anthropic's Claude Opus 4.5. The API pricing for DeepSeek V4 is 10-50 times cheaper than that of GPT-5.4 and Claude Opus 4.5; it is anticipated that DeepSeek V4 will be open-sourced under the Apache 2.0 license. DeepSeek V4 can run locally on dual RTX 4090 or single RTX 5090 setups. Additionally, DeepSeek introduced three groundbreaking innovations for DeepSeek V4: 1. Engram memory; 2. Manifold-constrained hyperconnection (mHC); 3. Sparse attention mechanism (DSA) and Lightning indexer. Furthermore, the DeepSeek official statement noted that due to strict U.S. export restrictions on high-end NVIDIA GPUs (such as the B300 and H200), DeepSeek has optimized V4 to primarily rely on domestically produced chips in China for inference. While initial training may still have utilized NVIDIA hardware (like H800s), the model has been highly optimized for Huawei's Ascend 950PR and Cambricon MLU chips.