Modal
Serverless cloud platform for running AI inference, training, and batch jobs with elastic GPU access.
Updated April 2026
Overview
- Website
- modal.com
- Segment
- Serverless Inference
Product overview
Modal provides serverless GPU compute for AI/ML workloads including inference, fine-tuning on multi-node clusters, batch processing at scale, sandboxes, and collaborative notebooks. It serves AI developers and teams at companies like Phonic, Orbital Materials, and Harvey.ai, enabling fast deployment without infrastructure management. Distinct for sub-second cold starts, Python-defined infra (no YAML), multi-cloud GPU pooling with no quotas, and scale-to-zero billing.
Revenue model
Pay-per-second usage: e.g., H100 GPU $0.001097/sec (~$3.95/hr), A100 80GB $0.000694/sec (~$2.50/hr), T4 $0.000164/sec (~$0.59/hr); CPU $0.0000131/core/sec, memory $0.00000222/GiB/sec. Plans: Starter (free $30/mo credits), Team ($250/mo + $100 credits), Enterprise (custom). No reserved/on-demand split; committed spend via marketplaces.
Moat
The search results do not contain specific information about the competitive moat of a company or product named 'Modal'. They provide general definitions and examples of competitive moats, such as proprietary technology, scale advantages, network effects, and cost advantages, applicable to businesses broadly.