昨天,Anthropic 发布了一款新模型并下调了价格。九十分钟后,OpenAI 也做了同样的事。
大多数商业 AI 应用都处在那个混乱的中间地带:多步骤工作流,既需要一个足够聪明的模型,又得是公司负担得起的价格。这是当今市场中最重要的部分,也是竞争最激烈的地方。降价就是证据。
6 月,Anthropic 凭借 Fable 5 将前沿价格定为每百万 token 10 美元和 50 美元。7 月,OpenAI 以 GPT-5.6 Sol 应战,定价为 5 美元和 30 美元,以每任务三分之一的成本实现了同等能力。
Opus 系列此前从未变动过。Opus 4.5、4、4.8 和 5 的定价均为每百万 token 5 美元和 25 美元。昨天的降价是头一回。
低端市场的情况更为极端。OpenAI 在 7 月将 Luna 降价 80%,昨天又再降 50%。
不只是闭源模型之间的竞争,开放模型同样会压低价格。在那些公开数据的网关上,AI 货架上的“通用款”承担了大部分 token 用量,其价格比闭源模型的混合价格低 86%。
大客户则通过微调追求更大的成本节省。Cursor 的 Composer 2 对开放权重基座 Kimi K2.5 进行了微调,相比其此前的自研模型将整体成本降低了 86%。Harvey 也采取了同样的做法,相比 Sonnet 5 将每单元格成本降低了 55%,同时得分还高于 Fable 5。
但市场的右侧长尾比几乎所有人预测的都要薄。Anthropic 的 Fable 5.1 是其能力最强、价格最贵的模型,在头十二天里仅占网关支出的 3.7%。其前代产品在 7 月恢复访问时峰值达到 13.2%,一个月后当 Opus 5 以半价发布时便跌至 4.9%。在大型企业客户中,前沿模型在 token 消耗中的占比从 8 月初的 53% 降至 9 月的 45%。
对智能的需求并非一座金字塔——顶端狭小富裕,为下方一切买单。它是一个中间厚实的正态分布。中间那部分按每美元能买到多少智能来消费。
智能的成本持续暴跌。企业对 AI 的需求却远没有变化得那么快。因此,满足固定需求的那一档能力会不断变得更便宜。
随着每美元所能换取的智能呈爆炸式增长,token 的分布可能会向大宗商品化转移。这一趋势是否会发生,将决定 AI 市场的经济格局。
Yesterday, Anthropic released a new model & cut its price. Ninety minutes later, OpenAI did the same.
Most business AI use is the messy middle: multi-step workflows that need a smart enough model at a price a company can afford. It is the most important part of the market today, & it is where the competition is fiercest. The price cuts are the evidence.
In June, Anthropic set the frontier price at $10 & $50 per million tokens with Fable 5. In July, OpenAI answered with GPT-5.6 Sol at $5 & $30, matching that capability at a third of the cost per task.
The Opus line had never moved. Opus 4.5, 4, 4.8 & 5 all listed at $5 & $25 per million tokens. Yesterday’s cut was the first.
It is even more extreme at the low end. OpenAI cut Luna by 80% in July, then cut it another 50% yesterday.
More than just closed source rivalry, open models deflate prices too. The generics on the AI grocery aisle run a majority of token volume on the gateways that publish data, at an 86% discount to the blended price of closed models.
Large customers pursue even greater savings with fine tuning. Cursor’s Composer 2 fine tuned Kimi K2.5, an open-weight base, cutting its overall cost 86% against its previous in-house model. Harvey did the same, cutting cost per cell 55% against Sonnet 5 while scoring higher than Fable 5.
But the right tail of the market is thinner than almost anyone forecast. Anthropic’s Fable 5.1, its most capable & most expensive model, commanded only 3.7% of gateway spending in its first twelve days. Its predecessor peaked at 13.2% when access was restored in July, then fell to 4.9% a month later when Opus 5 shipped at half the price. Among large corporate accounts, frontier models fell from 53% of token consumption in early August to 45% by September.
Demand for intelligence is not a pyramid with a small, wealthy peak paying for everything beneath it. It is a normal distribution with a fat middle. The middle buys intelligence per dollar.
Intelligence costs keep plummeting. What enterprises demand from AI does not change nearly as fast. So the tier that satisfies a fixed requirement keeps getting cheaper.
As intelligence per dollar explodes, the distribution of tokens may shift to commodity. Whether that happens will determine the economics of the AI market.