GPT-5.4 Pro hits 150 IQ on Mensa Norway test, OpenAI posts benchmark jump
OpenAI’s GPT-5.4 Pro has reportedly scored an IQ-equivalent 150 on a public Mensa Norway-style benchmark run by TrackingAI, up from OpenAI’s earlier o3 score of 136. The article frames this as a major step-change in AI capability signaling, arriving during a macro-heavy week focused on inflation, labor data, and central-bank messaging.
GPT-5.4 Pro’s reported 150 score is presented as outperforming 99.96% of the human population on that particular public yardstick. OpenAI’s release also claims GPT-5.4 is its most capable and efficient frontier model for professional work, with improved coding, tool use, and “computer use,” plus a context window claimed up to 1 million tokens. The piece notes OpenAI also reported gains on other benchmarks (GDPval, OSWorld-Verified), aligning directionally with the IQ-style result.
The article stresses that public IQ tests are imperfect and sensitive to test design, prompt structure, and possible familiarity effects. Still, it argues that a jump from 136 to 150—especially alongside progress in coding and long-horizon task handling—could influence enterprise budgeting, hiring, and workflow automation decisions.
For traders, the key relevance is indirect: faster AI capability improvement may shift expectations for future capex and software demand, potentially affecting sentiment around AI-adjacent tech themes rather than cryptocurrencies directly.
Neutral
该报道本质上是“AI 能力基准”新闻，而不是加密网络、监管或链上流动性的直接变化。GPT-5.4 Pro 在 Mensa Norway 风格测试中从 136 跳到 150，可能短期内强化市场对“AI 进展加速”的叙事，进而带动 AI 相关科技板块情绪；但它未提供会立即影响交易所资金、稳定币流动、链上需求或风险偏好的量化链路信号。因此对加密市场更可能是情绪层面的边际影响，整体偏中性。
短期方面，类似“模型能力跑分/基准领先”往往会带来媒体与资金的短暂关注，但通常难以立刻转化为可验证的加密资产现金流或协议层变化。长期方面，如果能力提升确实推动企业加速预算与自动化落地，可能间接利好与 AI 相关的基础设施与应用叙事；然而从历史经验看，这类影响通常需要更长的落地周期，且市场定价往往取决于后续产品收入、资本开支与采用率，而非单一基准分数本身。
综合来看：这条新闻更像“AI 产业预期更新”，而非“加密市场结构性驱动”，因此给出中性判断。