开放权重AI模型正在成为下一代AI生态的基础设施。据分析，中国模型在Hugging Face过去一年的模型下载量中占比达41%1，展现出开放权重领域的竞争格局。其中，Z.ai的GLM-5.2在SWE-bench Pro上得分62.1%，超过了GPT-4.5的58.6%1；同时，Moonshot公司承诺将在7月27日发布Kimi K3的权重1。
观察人士指出，美国不应限制中国开放权重模型的发展，而应采取积极的竞争策略。建议措施包括发布前沿级的美国开放权重模型、通过政府采购来推动开放市场的形成，以及建立独立的行业标准而非实施禁令1。分析认为，只有这样才能保持美国的AI领导地位，否则美国将面临失去生态主导权的风险1。
Open-weight AI models are increasingly functioning as foundational infrastructure for the artificial intelligence ecosystem, with significant implications for technological competition between the United States and China.1 Chinese models have captured 41% of downloads on Hugging Face over the past year, demonstrating their growing prominence in the global market.1
Performance metrics highlight the competitive pressure facing American developers. Z.ai's GLM-5.2 model achieved a score of 62.1% on the SWE-bench Pro benchmark, outperforming GPT-4.5's 58.6%.1 Chinese companies are accelerating their release timelines, with Moonshot committing to publish the weights for its Kimi K3 model on July 27.1
Rather than imposing restrictions on Chinese open-weight models, strategic analysts recommend the United States pursue alternative approaches to maintain technological leadership.1 Proposed measures include releasing frontier-level American open-weight models, leveraging government procurement to create competitive open markets, and establishing independent industry standards rather than relying on prohibitive policies.1
评论
还没有评论，欢迎留下第一条。