Lambda is a cloud provider offering GPU instances and clusters for training and inference, plus on-premises hardware, aimed at AI researchers and teams needing GPUs. It provides on-demand, reserved, and bare-metal GPU options billed per-GPU-hour with per-minute granularity, and there is no free tier. The platform does not offer serverless or spot instances. Founded in 2012, Lambda carries a SOC 2 attestation and serves both individual researchers and larger teams.
gpuh100on-demandbare-metal
Overview
- Founded
- 2012
- Pricing tier
- Mid
- Startup-friendly
- Yes
- Enterprise-ready
- Yes
GPU cloud
- Free tier
- No
- H100 $/GPU-hour
- 2.99
- Serverless
- No
- Per-second billing
- Yes
- Bare metal
- Yes
- Spot / interruptible
- No