Expired50% confidenceFactExact time
Running a 4-bit quantized 70B model on 1×A100 80GB costs about $1.09 per hour and can support about 98,000 in-flight tokens.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/13/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/27/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedRunning vLLM 0.27.1 official image to infer Qwen2.5-0.5B-Instruct (fp16) on an H100 SXM 80GB on vast.ai ($2.689 per hour, 20 vCPU), the cost per million output tokens is about $1.26 under a single-request serial load.70% similarUnverifiedFor high-quality models of 70B and above, local deployment typically requires multiple A100/H100-class GPUs, with hardware costs potentially reaching tens of thousands to hundreds of thousands of dollars.69% similarVerifiedRunning inference services on a 70B-parameter open-source model requires at least 2–4 A100/H100-class GPUs, with cloud rental costs of $5–20 per hour.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1019930API
curl https://kongchang.com/api/v1/knowledge/claims/1019930MCP
get_claim(id=1019930)