GPUs from $0.30/hr.
RTX 4070 SUPER (12GB) at $0.30/hr, RTX 4060 Ti (16GB) at $0.37/hr — tiered by VRAM, billed per second. Request an invite for 60 free minutes to try it out, then top up if you like it. Card (Toss) payment is still under review, so for now we top up your account manually upon request.
Cheaper than the big clouds
That's because we rent out consumer GPUs during their idle time. The big cloud providers charge roughly $0.5~$1/hr even for GPUs with similar VRAM, while we start at $0.30.
Our hourly price is cheaper than the big clouds' entry-level GPUs. But since we run a pool of personal PCs, there's no datacenter SLA and no large VRAM for big models — it's for light-to-medium workloads.
① Once invited, start right away with no sign-up and no foreign card · ② All guidance and troubleshooting in Korean (KakaoTalk) · ③ Per-second billing, so short jobs cost only what you use · ④ Image batch charges only for QC-passed outputs (rejects are free).
— For light-to-medium workloads, whether you can get this convenience at a price cheaper than the big clouds is the question this beta is trying to answer.
Pricing
GPU rental
Tiered by VRAM — the bigger the model you load, the more VRAM matters, so 16GB costs more. Billed per second, so short jobs cost only what you use.
- Once invited, start with a trial credit (60 free minutes)
- 70% paid out to providers, 30% to the platform
- Yields when the owner returns · session restarts if a node drops
Image batch
Background removal + 1024px standardization for product photos. Charged only for QC-passed outputs — we reject bad ones and don't charge for them.
- No sign-up, start right now (/image-batch)
- Processed by a real distributed GPU pool
GPT-2 inference
One text-generation call for the demo. Failures and timeouts are not charged.
- Per single API call
- ※ This is a small beta demo model
※ These are beta-period rates and may be adjusted as demand and cost data accumulate (with advance notice).
GPU lineup — take your pick
The listed price is the hourly rate for that GPU once it's in the pool. Right now only the 2 GPUs marked ● are actually rentable; for the rest we're still recruiting providers — we don't pretend to have what we don't.
| GPU | VRAM | Per hour (USD) | Status |
|---|---|---|---|
| RTX 4060 | 8GB | $0.22 | ○ Recruiting |
| RTX 3060 | 12GB | $0.24 | ○ Recruiting |
| RTX 3070 | 8GB | $0.24 | ○ Recruiting |
| RTX 4070 SUPER | 12GB | $0.30 | ● Available now |
| RTX 4070 Ti | 12GB | $0.33 | ○ Recruiting |
| RTX 5070 | 12GB | $0.34 | ○ Recruiting |
| RTX 4060 Ti | 16GB | $0.37 | ● Available now |
| RTX 4080 SUPER | 16GB | $0.46 | ○ Recruiting |
| RTX 3090 | 24GB | $0.50 | ○ Recruiting |
| RTX 5080 | 16GB | $0.52 | ○ Recruiting |
| RTX 4090 | 24GB | $0.66 | ○ Recruiting |
| RTX 5090 | 32GB | $0.95 | ○ Recruiting |
※ Beta indicative prices set by VRAM and generation. Actual rates may be adjusted based on the supply cost of the providers who join.
Is the GPU you want still recruiting?
Let us know and we'll prioritize recruiting a provider for that GPU, then email you as soon as it joins the pool.
Top up
This is a real paywall. Send a top-up request and we'll email you the payment steps (still manual during the beta). Automatic card payment (Toss) goes live the moment our merchant review passes.
Top-up request
Send the form below and we'll reply by email with the payment steps. Once confirmed, your credit is added.
· Automatic card payment (Toss Payments) is going through sole-proprietor, e-commerce, and merchant review. Until it's approved, payment runs as: top-up request → manual guidance.
· The current isolation level is process isolation — don't upload sensitive data, personal information, or proprietary models (isolation matrix).
· If the service is discontinued during the beta, any unused balance is fully refunded.