RunPod vs TensorDock
TensorDock is cheaper on 6 of the 8 GPUs both providers list with on-demand pricing, verified 2026-09-01. RunPod still wins on 2, its best case being the L40 at $0.82 vs $0.95. Platform facts come from each provider's own documentation.
Price per GPU, side by side (2026-09-01)
| GPU (on-demand $/hr) | RunPod | TensorDock | Price diff |
|---|---|---|---|
| RTX A4000 | $0.25 | $0.10 | -$0.15 (-60.0%) TensorDock |
| Tesla V100 | not offered | $0.17 | – |
| RTX 3090 | $0.50 | $0.20 | -$0.30 (-60.0%) TensorDock |
| RTX A5000 | $0.27 | not offered | – |
| RTX 4090 | $0.74 | $0.35 | -$0.39 (-52.7%) TensorDock |
| A40 | $0.44 | not offered | – |
| RTX A6000 | $0.53 | $0.45 | -$0.08 (-15.1%) TensorDock |
| L4 | $0.49 | not offered | – |
| RTX PRO 4500 | $0.72 | not offered | – |
| RTX 6000 Ada | $0.84 | $0.75 | -$0.09 (-10.7%) TensorDock |
| L40 | $0.82 | $0.95 | -$0.13 (-13.7%) RunPod |
| L40S | $0.99 | not offered | – |
| RTX 5090 | $0.99 | not offered | – |
| A100 | $1.39 | $1.50 | -$0.11 (-7.3%) RunPod |
| RTX PRO 6000 | $1.89 | not offered | – |
| H100 | $2.89 | $2.25 | -$0.64 (-22.1%) TensorDock |
| MI300X | $2.39 | not offered | – |
| H200 | $3.79 | not offered | – |
| B200 | $6.79 | not offered | – |
| B300 | $7.89 | not offered | – |
Cheapest verified on-demand rate per side from today's snapshot; highlighted = cheaper. Price diff is the absolute and percent saving on the cheaper side (named), relative to the higher price; – means one side does not offer that GPU on-demand. Spot/marketplace tiers shown on the individual pages.
Tables rank by price, lowest first. Commercial relationships never affect ordering.How we collect and rank prices.
Platform facts
| RunPod | TensorDock | |
|---|---|---|
| Billing increment | per-second | not published |
| Egress | free | not published |
| Storage /GB-mo | $0.10/GB/month volume disk; $0.07/GB/month network volume under 1TB | not published |
| Spot offered | not published | not published |
| Reserved offered | yes self-serve | not published |
| Notebook support | templates | not published |
From each provider's documentation, verified 2026-07-07; sources on the fees matrix and enterprise matrix.
FAQ
TensorDock is cheaper on 6 of the 8 GPUs both providers offer, verified 2026-09-01. The other is cheaper on the rest, so the right pick depends on the specific GPU you need.
The H100: $2.25 per GPU-hour on TensorDock vs $2.89 on RunPod, a difference of $0.64 per hour, verified 2026-09-01.
RunPod lists 11 GPUs with on-demand pricing that TensorDock does not: RTX A5000, A40, L4, RTX PRO 4500, L40S, RTX 5090 and 5 more. TensorDock lists 1 that RunPod does not: Tesla V100.
Prices render from the live dataset each build. More: the screener · price movers · methodology.