Google · Updated daily · last refresh 2026-10-01
TPU v6e pricing
What the Google TPU v6e (Trillium) costs right now: 1 tracked on-demand rates, the spread between them, and what the hourly number turns into when you own the card instead.
The headline on TPU v6e pricing is the spread: the same card costs 1.0x more at the top of the market than at the bottom. The floor comes from GCP (hyperscaler capacity), the ceiling from GCP, where you are paying for the ecosystem around the GPU as much as the silicon. Every rate below is an on-demand list price; commit to reserved or spot capacity and 30-70% comes off. Trillium: ~4.7× v5e compute per chip. Google's mainstream training silicon.
TPU v6e rental prices by provider
| Provider | $/GPU-hr | Notes |
|---|---|---|
| GCP | $2.70 | per chip-hour, us-east5 |
Put the price in context
TPU v6e specs and capability
What this card actually runs: memory, bandwidth, and the models that fit at each precision.
Convert to $/million tokens
Hourly pricing is the input. The estimator turns it into serving cost for any model.
TPU v5e vs TPU v6e
Head-to-head specs, pricing, and which models fit on one but not the other.
TPU v6e vs H100 SXM
Head-to-head specs, pricing, and which models fit on one but not the other.
All GPU pricing
The full table: every accelerator we track, cheapest to priciest, updated monthly.
Common questions
How much does a TPU v6e cost to rent?
As of 2026-09, TPU v6e on-demand rental runs from $2.70 per GPU-hour at Google Cloud to $2.70 at Google Cloud, with a median of $2.70 across the 1 providers we track. Reserved and spot pricing typically runs 30-70% below these list rates.
*Owned effective rate: 36-month amortization, 60% utilization, $0.08/kWh with cooling overhead; hosting and ops labor excluded. Full methodology.