Loading prices…
TPU v5e rental prices
Google TPU · TPU · 16 GB HBM2 · TPU · 2023
1 providers · 6 configurations · prices as of Nov 11, 2026, 05:01 UTC
Lowest on-demand
$1.20per GPU·h · ≈ $876/month
at Google CloudMedian on-demand
$1.20/GPU·h
across 1 providers
Lowest spot
$0.12/GPU·h
interruptible
FP16/BF16 cost
$6.09/PFLOP·h
#61 of 72 GPUs for value
Providers
1
6 configurations
Price history
Daily TPU v5e prices per GPU-hour. The shaded period is reconstructed from public archives: the version history of SkyPilot's price catalogs, Internet Archive copies of providers' pricing pages and the open gpu-price-feed project (fewer providers than live data). After that, PetaPrice's own tracking.
Lowest on-demandMedian on-demandLowest spot
Price by provider
Cheapest On-demand TPU v5e offer at each provider, per GPU-hour. Shorter and greener is cheaper.
Google Cloud🇺🇸$1.20best
Thin vertical line = median across providers ($1.20).
Every provider we track
Pricing
Size
Host
Provider HQ
1 providers · 6 configurations
Sort
- Google CloudOn-demandSpot🇺🇸 HQ in United Statesvia SkyPilot catalogfrom · On-demandCheapest$1.20/GPU·hCheapest
What can the TPU v5e run?
Memory for the weights plus one 32K-token sequence of context, how many GPUs hold it, and what that replica costs at the cheapest in-stock on-demand rate.
| Model | Params | Memory | GPUs | $/hr | $/mo |
|---|---|---|---|---|---|
gpt-oss-20b OpenAI · 20.9B | 20.9B · 3.6B active | 20 GiB | 2 | $2.40 at Google Cloud | $1,752 |
Gemma 3 27B Google · 27.4B | 27.4B | 28 GiB | 2 | $2.40 at Google Cloud | $1,752 |
Qwen3-32B Alibaba · 32.8B | 32.8B | 39 GiB | 4 | $4.80 at Google Cloud | $3,504 |
Llama 3.3 70B Instruct Meta · 70.6B | 70.6B | 76 GiB | 8 | $9.60 at Google Cloud | $7,008 |
gpt-oss-120b OpenAI · 117B | 117B · 5.1B active | 110 GiB | 8 | $9.60 at Google Cloud | $7,008 |
Qwen3-235B-A22B Alibaba · 235B | 235B · 22B active | 225 GiB | 16 | $19.20 at Google Cloud | $14,016 |
DeepSeek-R1 DeepSeek · 684B | 684B · 37B active | 640 GiB | 64 | $76.80 at Google Cloud | $56,064 |
Kimi K3 Moonshot AI · 2.78T | 2.78T · 104B active | 2.53 TiB | 64+ | — | — |
Assumes 90% of each GPU's memory is usable and rounds up to 1, 2, 4, 8… GPUs for tensor parallelism (up to 64). $/hr is the GPU count times the cheapest in-stock on-demand rate per GPU; check that the provider rents instances that large. Serving many requests at once needs more memory for context.
Alternatives to the TPU v5e
GPUs with similar compute and memory, and their lowest on-demand price right now.
TPU v5e pricing questions
How much does it cost to rent an TPU v5e?
On-demand TPU v5e prices start at $1.20 per GPU-hour (Google Cloud). The median across 1 providers is $1.20, and the most expensive is $1.20. Running one GPU around the clock costs about $876 per month at the lowest price.
Is there cheaper spot or interruptible TPU v5e capacity?
Yes. Spot (interruptible) TPU v5e capacity starts at $0.12 per GPU-hour, 90% below the cheapest on-demand price. Spot instances can be preempted at any time, so use them for checkpointed training or batch jobs.
How much compute do you get per dollar with the TPU v5e?
At $1.20 per hour and 0.197 PF of dense FP16/BF16, the TPU v5e costs $6.09 per PFLOP-hour. That ranks 61 of 72 GPUs with on-demand prices on PetaPrice. Figures are datasheet peaks; real utilization is typically 30–60%.
What are the TPU v5e's key specs?
16 GB HBM2 with 819 GB/s of memory bandwidth, TPU architecture (2023), TPU form factor. Dense peak: 0.197 PF FP16 / BF16, 0.394 PF INT8.
How current are these prices?
Prices were collected on Nov 11, 2026, 05:01 UTC from provider APIs and public catalogs and refresh about every 2 hours. They are indicative list prices; always confirm on the provider's website before booking.