Novita AI
thecontinuitylayer.com/tools/gpu-hosting/novita-aiDevelopers who want both an open-model API and rentable GPUs from one account
GPU cloud · Novita AI
Novita AI offers a per-token API for many open-weight models alongside GPU instances, spot instances, serverless GPUs and dedicated model deployments. The GPU side lets you start a container on a chosen card and pay by the hour.
It is handy if you want to start on the API and move to your own GPU later without changing provider. Its published GPU prices vary with market conditions and differ across its own pages, so check the console for live rates.
Specs
| Price per hour | RTX 4090 $0.67/hr on demand; spot instances about half price |
|---|---|
| GPUs | Consumer and data-center GPUs such as RTX 4090, RTX 5090, H100 and H200, plus a per-token model API |
| RTX 4090 | $0.67/hr on demand (varies with market) |
| Spot discount | Roughly 50% below on demand |
| GPU types | RTX 4090, RTX 5090, H100, H200 and others |
| Products | GPU instances, serverless GPUs, dedicated deployments, model API |
Good for
- Mixing API calls with self-hosted GPUs
- Cheap spot runs on consumer cards
Not for
- Buyers who need stable, clearly published rates
- Large multi-node clusters
See also
- GPU Droplets · Developers already on DigitalOcean who want a GPU server next to their app
- Hyperbolic · Developers who want on-demand data-center GPUs without a contract
- Modal · Python developers who want to serve an open model as code, paying only while it runs
We last reviewed Novita AI on 31 October 2026. Prices and specs change; check novita.ai before you buy or sign up. Figures we could not confirm on the vendor's own page read “not published” or “varies”.
Vendor source: https://novita.ai/gpu-instance/gpu/rtx-4090