GPU RENTAL ON SOLANA

Rent compute by the hour.
Settle on-chain.

Pick a card, pick your hours — pay with SOL or the dolphy token.
Pods spin up in seconds. No sign-up, no middleman.

Rent a GPURent a GPU
QUOTE/api/quote200 OK
{
    "gpu": "H100 · 80GB",
    "hours": 6,
    "price_sol": "0.42",
    "price_token": "3.57",
    "discount": "15%",
    "expires_in": 60
}
Tencent
ByteDance
NetEase Games
Virtual Flow
Bambu Lab
HTC
Civitai
Makeronline
stability.ai
Layer AI
Nilo
Style 3D

dolphydolphy

From wallet to running pod in three steps.

Connect your wallet

Phantom, Solflare or an embedded wallet — social login works too. No accounts, no emails, nothing to install.

WALLET CONNECTED — DEVNET
$ phantom connect ✓ 9WzDX…3b2F — ready to pay (devnet)Try it in the marketplace
Try it in the marketplace

WHY DOLPHY

Bare-metal GPUs,
on-chain simplicity.

No sign-up, no KYC, no monthly plans. Connect a wallet, lock a quote, pay — the market does the rest.

01 — On-chain payments

Pay with SOL or token.

Every rental is a single on-chain transfer to the treasury — verifiable by anyone. Hold the dolphy token and get 15% off every hour.

One tx. Zero paperwork.

02 — Instant provisioning

Paid means running.

Payment verified → RunPod provisions your pod in seconds. SSH in immediately — no queues, no humans in the loop.

Seconds, not support tickets.

03 — Open market

Idle GPUs start earning.

Capacity flows to where it's needed — prices set by supply and demand, not by sales calls. Renters pay market rates, GPU owners monetize idle hardware.

Supply meets demand.

Your keys, your compute.

MARKETPLACE — GPU BY THE HOUR

Rent GPUs with SOL — 15% off with token

Pick a card, pick hours, pay on-chain. RunPod provisions in seconds. Mock mode until live keys.

MOCK — no RunPod key — Any SPL token for testing (devnet). Mainnet mint later.

RTX 4090

24GB · GDDR6X

Available

Best for inference, 3D, gaming. 24GB VRAM.

SOL /hour

0.002300

SOL

-15%

TOKEN /hour

0.001950

TOKEN

Rent →Starting at

RTX 6000 Ada

48GB · GDDR6

Available

48GB VRAM for large models, rendering.

SOL /hour

0.003500

SOL

-15%

TOKEN /hour

0.002970

TOKEN

Rent →Starting at

L40S

48GB · GDDR6

Available

Data-center efficiency, 48GB, for inference at scale.

SOL /hour

0.004100

SOL

-15%

TOKEN /hour

0.003480

TOKEN

Rent →Starting at

A100 40GB

40GB · HBM2

Available

Training and large-batch inference.

SOL /hour

0.005600

SOL

-15%

TOKEN /hour

0.004760

TOKEN

Rent →Starting at

A100 80GB

80GB · HBM2e

Available

80GB for biggest models, long context.

SOL /hour

0.007300

SOL

-15%

TOKEN /hour

0.006200

TOKEN

Rent →Starting at

H100 80GB

80GB · HBM3

Limited

Flagship Hopper, fastest for LLM training.

SOL /hour

0.01300

SOL

-15%

TOKEN /hour

0.01105

TOKEN

Rent →Starting at
View full marketplace →

GPUs online

4,821

available right now

Avg price

0.74/hr

market rate, USD

Rentals settled

12,908

100% on-chain

With token

−15%

every hour, every GPU

Prices are set by the market. Payments are set by you.

OPEN MARKETPLACE →