Unverified50% confidenceFactExact time
Hyperscale cloud vendors rent out computing power by the number of virtual CPU cores or instance hours, while in the AI era billing by Token has begun for inference services, with revenue directly tied to Token throughput.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/14/2026
First Seen
Valid until: 1/12/2027
Sources
Related Entities
Related Claims
UnverifiedIn the AI field, cloud vendors building their own AI data centers is a typical heavy-capital-expenditure model, requiring upfront investment in the tens of billions of dollars, while the corresponding AI revenue is realized gradually by usage; when CapEx growth far exceeds revenue growth, ROI may be deteriorating even if absolute revenue is growing.77% similarUnverifiedAs cloud vendors expand AI compute supply and H100/H200 and other chips enter mass production, compute advantages are shifting from differentiated assets to infrastructure.74% similarUnverifiedTogether AI's inference cloud service adopts a per-token billing model and provides inference capabilities through optimizations such as quantization, Flash Attention, and continuous batching.73% similarUnverified微软CEO萨提亚·纳德拉认为企业使用云端AI是在'为智能付费两次'，一次用金钱，另一次用为了让智能有用而必须泄露的专有知识73% similarUnverifiedHyperscale cloud providers such as AWS, Azure, and Google Cloud have invested tens of billions of dollars in AI infrastructure.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1032307API
curl https://kongchang.com/api/v1/knowledge/claims/1032307MCP
get_claim(id=1032307)