Welcome to Opensource AI Cloud
GPU compute
in minutes
Run AI workloads on a real GB10 Blackwell GPU. This tour shows you everything — from account creation to getting your results back.
What you can run
AI model inference
Model fine-tuning
CUDA workloads
Data processing
skip to account creation
Step 1 of 4 — Account
Create your
free account
Register with just your email and password. No credit card needed — you get 15 minutes of GPU compute completely free.
Register via API or web
$ curl -X POST .../backend/auth/register \
Step 2 of 4 — Submit
Submit your
GPU job
Send any bash command to the jobs API. It runs inside an isolated Docker container with full access to the GB10 Blackwell — 128 GB unified memory.
Submit a job
$ curl -X POST .../backend/jobs/submit \
-H "Authorization: Bearer TOKEN" \
-d '{"command":"python3 train.py --epochs 10"}'
{"id": 42, "status": "RUNNING"} ✓
Step 3 of 4 — Results
Get your
results back
Poll the status endpoint. When complete, the full output of your job is returned — stdout, stderr, and exit code all captured.
Job lifecycle
PENDING
→
RUNNING
→
COMPLETED
Step 4 of 4 — Billing
Pay only for
what you use
Your first 15 minutes are free. After that add a card and pay per minute — no subscriptions, no minimums.
Pricing
Free trialfree 15 mins
Inference£0.10 / min
Training£0.25 / min
Dedicated£1.50 / hr