Modal
Serverless cloud for running AI inference, training and batch workloads
About Modal
Modal is a cloud platform where developers run Python functions on GPUs for inference, training, batch jobs and sandboxed code execution. It bills by the second and scales to zero when idle.
Common use cases
- LLM inference
- Batch processing
- Model fine-tuning
- Agent sandboxes
Products and services
Inference
Serving LLMs and image, video and audio models.
Training
Fine-tuning and RL on GPUs.
Sandboxes
Isolated environments for agents and untrusted code.
Functions and Batch
Compute-heavy functions and async batch jobs.
Key features
- Serverless GPUs
- Per-second billing
- Autoscaling to zero
- Sandboxes
- Multi-cloud capacity
- Observability
Integrations
- PyTorch
- Hugging Face
- vLLM
- AWS
- OpenAI-compatible APIs
Modal alternatives
FAQs
How is Modal priced?
Per second of usage, with a monthly free credit.
Does it scale to zero?
Yes, functions scale to zero when idle.
What workloads?
Inference, training, batch jobs and sandboxes.
Work at Modal? Request updates to this profile. Update this profile
Compare with similar providers
All ML Operations providersAnyscaleManaged platform for running Ray-based AI workloads at scale
Blend360Data science, analytics and AI services firm
Edge ImpulsePlatform for building and deploying machine learning on edge devices
Hugging FaceHub for open source models and datasets with hosted inference and enterprise tools
IguazioMLOps and generative AI operations platform built on the open source MLRun
LablupBackend.AI platform for GPU and AI infrastructure management
LaminiPlatform to tune, evaluate and deploy LLMs on private infrastructure
MakinarocksIndustry-specific AI operating system and AI agents for manufacturing and defense