Expired60% confidenceFactTime unknown
On an NVIDIA H100 GPU, running a model of approximately 200 billion parameters (Opus scale), the cost of generating 1,000 tokens in a single inference is approximately $0.01–$0.03.
1
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedFor a 70-billion-parameter model (Sonnet scale) optimized through distillation and architectural improvements, the cost of generating 1,000 tokens in a single inference is approximately $0.003–$0.005.78% similarUnverifiedA single NVIDIA H100 GPU costs around $30,000-$40,000, and running a large language model typically requires dozens or even hundreds of GPUs working in parallel.77% similarUnverified一块H100 GPU租用成本约每小时2-4美元，单卡每秒可生成约1000-3000个输出Token77% similarUnverifiedFor a cluster of 10,000 H100s, the GPU hardware cost alone could reach $300 million to $400 million.77% similarUnverifiedThe purchase cost of a single A100 GPU exceeds $10,000, and its utilization rate directly determines service profitability.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/35036API
curl https://kongchang.com/api/v1/knowledge/claims/35036MCP
get_claim(id=35036)