Analysis: The gap between the open source big model and the closed source frontier in China and the United States is only 3 to 6 months, accelerating the global trend of replacement at a very low cost According to Beating monitoring, aggregation service provider OpenRouter has revealed that the performance gap between open source models and closed source cutting-edge models has stabilized at 3 to 6 months. In the past 18 months, Frontier Closed Source Laboratory has not been able to expand its position as expected, while open source forces represented by new players from China and the United States are accelerating the replacement of closed source models with extremely high cost-effectiveness. DeepSeek V4 Flash became the preferred replacement just two months after its release. The DeepSeek V4 Flash with 284 billion parameters achieved a score of 79.0% in SWE bench Verified evaluation, approaching the GPT-5.5 level. The official first party input/output pricing is only $0.14/0.28 per million tokens, and the output cost is about 150 times cheaper than GPT-5.5. Even with the Western cloud hosting premium that does not retain data for training, the actual cost is only around 1.3% of that of closed source frontier models. In addition to its price advantage, the GLM 5.2 released by Zhipu in June 2026 ranks first in the Artificial Analysis Open Source Weighted Intelligence Index and is comparable to the GPT-5.5 level in real agent evaluations, becoming an alternative solution for long-range programming planning. However, GLM 5.2 consumes more tokens during deep thinking, and enterprise deployment needs to balance output costs. The multimodal open-source model MiniMax M3, with its innovative MSA sparse attention architecture, provides native image and video long context processing capabilities at a lower token price, making it a strong open-source competitor to Gemini Flash. Meanwhile, Nvidia Nemotron 3 Ultra, based on the Mamba-2 hybrid architecture, has become the strongest domestic open source force in the United States, aiming to drive market demand for Nvidia's hardware and microservices ecosystem through an open ecosystem. OpenRouter emphasizes that although cutting-edge closed source models will eventually move forward, the cost of tokens with fixed intelligence levels will continue to decrease, providing significant cost optimization space for enterprises. [Original link]