Trader sentiment on A100 GPU rental prices at end-November centers on abundant secondary-market supply via platforms like Vast.ai and RunPod, where on-demand rates currently range from roughly $0.40–$1.80 per GPU-hour with medians near $1.50–$1.80, offset by steady demand for cost-effective inference and smaller-model training amid newer Blackwell and Hopper availability. Recent 90-day data show modest median declines or flat trends as older Ampere cards remain plentiful, yet any surge in large language model workloads or developer conferences could tighten availability and lift prices toward the $1.00–$1.25 band. Key swing factors include spot versus reserved contract shifts, hyperscaler capacity releases, and whether AI inference demand outpaces the expanding used-GPU inventory before year-end.