GPU Rental Market Report: November 10, 2026

Ten GPU models changed price by more than 1 percent compared to November 9, 2026. The RTX 4070S Ti (16 GB) on vast.ai rose from $0.0944 to $0.203 per hour, a 115.0 percent increase. The RTX A4500 (20 GB) on runpod increased 74.5 percent, moving from $0.1089 to $0.19 per hour. The RTX PRO 4500 (32 GB) on runpod also rose 74.5 percent, from $0.4126 to $0.72 per hour.

Conversely, the GTX 1660 (6 GB) on vast.ai dropped 46.4 percent to $0.3634 per hour, down from $0.6778. The A40 (45 GB) on vast.ai fell 43.0 percent to $0.3361 per hour, from $0.59. The Q RTX 6000 (24 GB) on vast.ai decreased 34.8 percent to $0.1318 per hour, and the RTX 4070 Ti (12 GB) on vast.ai dropped 34.5 percent to $0.1244 per hour.

Other daily changes include the RTX 6000Ada (48 GB) on vast.ai rising 38.3 percent to $0.6484 per hour, the L40 (48 GB) on runpod increasing 35.8 percent to $0.69 per hour, and the RTX 2060 (12 GB) on vast.ai climbing 32.2 percent to $0.0822 per hour. Forty-three chips remained unchanged or moved within 1 percent of their previous price.

Compared to November 3, 2026, the L40 (48 GB) on runpod saw the largest weekly increase, rising 89.4 percent from $0.3644 to $0.69 per hour. The RTX PRO 4500 (32 GB) on runpod increased 81.0 percent over the week, from $0.3978 to $0.72 per hour.

The RTX 5000Ada (32 GB) on vast.ai experienced the largest weekly decline, falling 79.8 percent from $2.1211 to $0.4281 per hour. The GTX 1660 (6 GB) on vast.ai dropped 46.4 percent to $0.3634 per hour, and the L4 (22 GB) on vast.ai decreased 44.8 percent to $0.2516 per hour.

The A100 SXM4 (40 GB) on vast.ai rose 44.2 percent to $0.8022 per hour, while the A40 (45 GB) on vast.ai fell 43.0 percent to $0.3361 per hour. The Tesla P100 (16 GB) on vast.ai increased 39.7 percent to $0.0668 per hour. Thirty-nine chips were unchanged or within 1 percent of their price from a week ago.

Four new GPU models appeared in the tracking data since yesterday: Q RTX 5000, RTX 4000Ada, RTX 4070 laptop, and RTX 4500Ada. Three models disappeared from the listings: A10, RTX 4090D, and RTX 6000D.

Over the past week, six new models were added to the tracked set: A16, Q RTX 4000, Q RTX 5000, RTX 2070, RTX 4000Ada, and RTX 4070 laptop. Seven models were removed from the market data: GB300, GTX 1650, RTX 3070 Ti, RTX 5880Ada, RTX 6000D, Tesla P4, and Tesla V100S.

For GPUs with at least 24 GB of VRAM, the cheapest option is the Tesla V100 (32 GB) on vast.ai at $0.0689 per hour, with 19 offers available. The same card is the cheapest for the 32 GB tier at the same price.

In the 48 GB tier, the RTX A6000 on runpod is the lowest priced at $0.33 per hour. For the 80 GB and 96 GB tiers, the GB10 (119 GB) on vast.ai is the cheapest at $0.3756 per hour, with 9 offers. The MI300X (192 GB) on runpod is the cheapest for the 141 GB tier at $2.39 per hour.

The RTX 4090 on vast.ai has the highest number of offers, with 33 listings available at two price points: $0.1356 per hour and $0.4563 per hour. The CMP 170HX on vast.ai follows with 32 offers at $0.2418 per hour and $0.4012 per hour. The RTX 3080 on vast.ai has 30 offers at $0.0928 per hour.