AIHOT 于 2026-08-18 收录了“Anthropic 在 Vercel AI Gateway 的 token 成本是其他厂商平均的 4.4 倍”这一公开动态。以下先呈现从来源页面抓取的正文，再给出 AIHOT 摘要与 TopoReduce 编辑解读。
PUBLIC SOURCE CONTENT
已抓取公开正文公开原文内容
Anthropic's per-token cost runs 4.4 times the average on Vercel, and developers keep paying
Anthropic's per-token cost runs 4.4 times the average on Vercel, and developers keep paying
Matthias Bastian
View the LinkedIn Profile of Matthias Bastian
Aug 18, 2026
Anthropic accounts for 65.1 percent of spending on Vercel's AI Gateway but only 30 percent of tokens. Its per-token cost runs 4.4 times the average of every other provider, meaning the company earns far more per use than its competitors.
Anthropic dominated July spending on Vercel's AI Gateway, while xAI and Moonshot grabbed a visible share of open-weight model spend for the first time. | Image: Vercel
Fable 5 quickly climbed to 13.2 percent of Gateway spending, second only to Opus 4.8. Nine out of ten Fable teams in July were new customers, which Vercel reads as a sign that the model is attracting fresh users and that willingness to pay for premium models remains strong. Payments company Ramp had previously flagged relatively low Fable usage as evidence the model might be priced too high. Both companies only see a slice of the full picture.
Token volume on Vercel rose 59 percent overall, and spending grew 37 percent. Average token prices fell 13.6 percent, though, as companies shift toward cheaper tiers.Ad
Ad
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
Subscribe now
Source: Vercel
Matthias Bastian
View the LinkedIn Profile of Matthias Bastian
Aug 18, 2026
Anthropic accounts for 65.1 percent of spending on Vercel's AI Gateway but only 30 percent of tokens. Its per-token cost runs 4.4 times the average of every other provider, meaning the company earns far more per use than its competitors.
Anthropic dominated July spending on Vercel's AI Gateway, while xAI and Moonshot grabbed a visible share of open-weight model spend for the first time. | Image: Vercel
Fable 5 quickly climbed to 13.2 percent of Gateway spending, second only to Opus 4.8. Nine out of ten Fable teams in July were new customers, which Vercel reads as a sign that the model is attracting fresh users and that willingness to pay for premium models remains strong. Payments company Ramp had previously flagged relatively low Fable usage as evidence the model might be priced too high. Both companies only see a slice of the full picture.
Token volume on Vercel rose 59 percent overall, and spending grew 37 percent. Average token prices fell 13.6 percent, though, as companies shift toward cheaper tiers.Ad
Ad
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
Subscribe now
Source: Vercel
AIHOT 摘要
Anthropic 占 Vercel AI Gateway 支出的 65.1%，但仅占 token 总量的 30%，单 token 成本是其他厂商平均水平的 4.4 倍。7 月 Gateway 总 token 量环比增长 59%，支出增长 37%，平均 token 价格下降 13.6%。Fable 5 攀升至支出的 13.2%，其 7 月团队中九成为新客户。
为什么值得关注
同一模型在不同平台显示相反信号，Ramp 看到的低使用与 Vercel 看到的高新客占比并存，提示单一数据源可能误判定价效果。
工程化解读
从 TopoReduce 的工程视角看，这条信息属于“基础设施与部署”主题。它的价值不只在于一个新产品或新观点本身，还在于说明 AI 系统正在如何影响模型接入、智能体协作、研发流程、基础设施和团队决策。实际采用前，应结合原文确认版本、适用范围、价格和运行条件。
- 发布时间：2026-08-18；AIHOT 分类：基础设施与部署。
- AIHOT 标签：
- AIHOT 判断：同一模型在不同平台显示相反信号，Ramp 看到的低使用与 Vercel 看到的高新客占比并存，提示单一数据源可能误判定价效果。
- AIHOT 评分：55；评分用于站内排序，不等同于独立评测结论。
TopoReduce 编辑观察
当 AI 动态进入真实生产环境，团队需要同时关注能力边界、数据来源、调用成本、权限控制和可回滚性。把单条新闻放回完整工程链路中阅读，比只看标题更有助于判断它是否适合自己的产品和工作流。