Cheapest
Lowest true cost per GPU-hour, in stock or stock unknown, newest reading per provider
Akash Network1× · 00:20 UTC$0.06

gcube1× · Korea$0.07
- 4
Cherry Servers1× · Lithuania$0.16
Rented today by 4 providers with a current on-demand offer, from $0.04 to $0.16 per GPU-hour, median $0.07. Every number below links to the page it was read from, with its read date.
Specifications: manufacturer figures.
Cheapest, median and spread among current offers. Medians count one floor per marketplace and need at least three offers.
Today's podium · on-demand · readings of the last 45 days · reference Nov 2, 2026 00:32 UTCCheapest true cost, best throughput per dollar, highest Trust Score. One provider per step, ten places, every row dated.







Current offers ranked by true cost, each with its read date. Older readings stay listed below, in grey, and never enter a ranking.
| Provider | Configuration | Region | Stock | List | True cost | Read | |
|---|---|---|---|---|---|---|---|
| 1× Tesla P44 vCPU · 32 GB RAM · 159.666 GB disk | US | Unavailable | $0.02 | $0.02 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P44 vCPU · 32 GB RAM · 159.453 GB disk | US | Unavailable | $0.02 | $0.02 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P44 vCPU · 32 GB RAM · 159.418 GB disk | Canada | Unavailable | $0.02 | $0.02 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P44 vCPU · 32 GB RAM · 159.59 GB disk | Australia | Unavailable | $0.02 | $0.02 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P44 vCPU · 32 GB RAM · 123.375 GB disk | Singapore | Unavailable | $0.02 | $0.02 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P44 vCPU · 32 GB RAM · 138.703 GB disk | Singapore | Unavailable | $0.02 | $0.02 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P44 vCPU · 32 GB RAM · 138.608 GB disk | US | Unavailable | $0.03 | $0.03 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P42 vCPU · 24 GB RAM · 155.135 GB disk | US | Unavailable | $0.03 | $0.03 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P412 vCPU · 16 GB RAM | Romania | Unavailable | $0.03 | $0.03 | 00:03 UTCfresh | Receipt | |
| 1× Tesla P46 vCPU · 15.6 GB RAM · 164 GB disk | Poland | In stock | $0.04 | $0.04 | 00:02 UTCfresh | Receipt | |
| 1× Tesla P4 | Akash Network providers (decentralised) | In stock | $0.06 | $0.06 | 00:20 UTCfresh | Receipt | |
| 1× Tesla P412 vCPU · 16 GB RAM · 512 GB disk | Korea | Stock unknown | $0.07 | $0.07 | 00:24 UTCfresh | Receipt | |
| 1× Tesla P4 | Lithuania | In stock | $0.16 | $0.16 | 00:19 UTCfresh | Receipt | |
| 1× Tesla P4 | Australia +19 | Stock unknown | $0.60 | $0.60 | 00:19 UTCfresh | Receipt | |
| 1× Tesla P48 vCPU · 32 GB RAM | Thailand | Stock unknown | $1.26 | $1.26 | 00:24 UTCfresh | Receipt | |
| 4× Tesla P432 vCPU · 128 GB RAM | HK | Stock unknown | $1.31 | $1.31 | 00:24 UTCfresh | Receipt | |
| 1× Tesla P48 vCPU · 32 GB RAM | HK | Stock unknown | $1.32 | $1.32 | 00:24 UTCfresh | Receipt | |
| 2× Tesla P416 vCPU · 64 GB RAM | HK | Stock unknown | $1.32 | $1.32 | 00:24 UTCfresh | Receipt |
The same hour, divided by what the card gives you; then the smallest real configuration for the most downloaded open models.
| Unit | At the cheapest | At the median |
|---|---|---|
| per GB of VRAM · hour8 GB | $0.005 | $0.008 |
| per TB/s of bandwidth · hour0.19 TB/s | $0.206 | $0.351 |
| Model | Needs | Smallest config | $ / month |
|---|---|---|---|
| all MiniLM L6 v2Sentence Transformers | 1 GBFP8 | 1× Tesla P4at Vast.ai | $29 |
| paraphrase multilingual MiniLM L12 v2Sentence Transformers | 1 GBFP8 | 1× Tesla P4at Vast.ai | $29 |
| bge-m3BAAI | 1 GBFP8 | 1× Tesla P4at Vast.ai | $29 |
| Qwen3 0.6BQwen | 1 GBFP8 | 1× Tesla P4at Vast.ai | $29 |
| all-mpnet-base-v2Sentence Transformers | 1 GBFP8 | 1× Tesla P4at Vast.ai | $29 |
| bge-reranker-v2-m3BAAI | 1 GBFP8 | 1× Tesla P4at Vast.ai | $29 |
Today the Tesla P4 rents from $0.04 per GPU-hour on-demand, with a median of $0.07 across 4 providers, each counted once at its cheapest current offer (4 current offers in all, readings of the last 45 days). Prices depend on the provider, the number of GPUs in the machine, the region and whether you rent on-demand, spot or reserved; true cost adds the mandatory add-ons each provider publishes.
6 providers have a current Tesla P4 offer in stock. Cheapest first: Vast Ai ($0.04), Akash Network ($0.06), Gcube ($0.07), Cherry Servers ($0.16), Google Cloud ($0.60), Huawei Cloud ($1.26). Older readings from other providers stay listed on this page, in grey, and are never ranked.
8 GB of memory at 0.19 TB/s. At FP8 that fits all MiniLM L6 v2 (1 GB), paraphrase multilingual MiniLM L12 v2 (1 GB), bge-m3 (1 GB), Qwen3 0.6B (1 GB), all-mpnet-base-v2 (1 GB) on a single card; larger models need two to eight GPUs or a different card.
The lowest true cost among eligible offers is $0.04 per GPU-hour at Vast.ai for a 1-GPU machine in Poland, read within the ranking window. Offers marked unavailable, older than 45 days, superseded by a newer reading, resold or outside the outlier band are listed but not ranked.
Spot (interruptible) capacity starts at $0.03 per GPU-hour; reserved terms start at $0.13. Spot suits jobs that checkpoint; reserved suits machines that run all month.