Koyeb
Serverless GPUs with per-second billing and global deployment
Last reviewed May 8, 2026
Koyeb is a serverless platform offering on-demand NVIDIA GPUs from RTX-class up through B200 with per-second billing, scale-to-zero and global edge deployment. Koyeb is joining Mistral AI to build the future of AI infrastructure.
Available GPUs
Hourly on-demand pricing. Click column headers to sort.
Prices last updated: July 31, 2026
Pros & Cons
Advantages
- Per-second billing across the full GPU range
- Multi-GPU configurations (2x, 4x, 8x) available on-demand
- Serverless model removes idle-instance costs
Limitations
- Smaller per-cluster scale than dedicated training neoclouds
- Some GPU SKUs require requesting access
- Long-running fixed-instance training is not the primary use case
Key Features
Serverless GPU Instances
Per-second billing with scale-to-zero across the GPU catalog
Wide GPU Range
From RTX-4000-SFF-ADA and L4 through L40S, A100, H100, H200 and B200
Global Edge Deployment
Deploy applications close to users across multiple regions from a single push
Pricing Options
| Option | Details |
|---|---|
| Per-Second GPU Billing | Charged per second of GPU runtime, with scale-to-zero when idle |
| Multi-GPU Configurations | Pre-configured 2x, 4x and 8x GPU instances for larger workloads |
Availability & Support
Regions
Global edge presence across North America, Europe and Asia
Support
Documentation, community forum and paid support tiers
Getting Started
- 1
Create an account
Sign up via the Koyeb console using email or GitHub
- 2
Pick a GPU instance
Select an instance type from the GPU catalog and configure your service
- 3
Deploy from Git or container
Push from a GitHub repo or Docker image and Koyeb handles the rest
Compare Providers
Find the best prices for the same GPUs and models from other providers
