Modal

by Modal Labs

Serverless platform for running ML inference, training, fine-tuning and data pipelines on GPUs and CPUs, billed per second.

Visit website

MLOps Platforms

Starting priceStarter $0 plus compute with $30/month free credits; Team $250/month plus compute

Pricing modelusage-based

DeploymentCloud-hosted serverless platform

Free trialNo

Company sizeStartups to enterprises

About

Serverless platform for running ML inference, training, fine-tuning and data pipelines on GPUs and CPUs, billed per second.

Modal is a serverless compute platform where teams deploy any containerizable Python application and have it autoscale instantly with request volume. It is well suited to compute-intensive workloads such as ML inference at scale, model training and fine-tuning, data pipelines and batch job queues.

The platform includes custom images, secrets, cron jobs, distributed queues and dicts, real-time metrics and logs, region selection, deployment rollbacks and environment-level budgets. Enterprise plans add audit logs, SSO and HIPAA compatibility. Billing is per second of compute, with nothing charged for idle time and monthly free credits on every plan.

Key features

Serverless GPU and CPU compute; Instant autoscaling; Custom container images; Cron jobs and distributed queues; Real-time metrics and logs; Deployment rollbacks; Environment-level budgets; SOC 2 compliance, SSO and RBAC

Product details

Vendor

Modal Labs

Pricing

Starter $0 plus compute with $30/month free credits; Team $250/month plus computeNo free trial

Pricing model: usage-based

Plans

Starter: $0 plus compute, $30 monthly free credits, 3 seats. Team: $250 per month plus compute, $100 monthly free credits, unlimited seats. Enterprise: custom, volume discounts.

Visit website for full pricing details

Similar software

Amazon SageMaker

by Amazon Web Services

Amazon SageMaker is AWS's unified platform for data, analytics and AI, covering model building, training, deployment and governance.

Pay as you go, AWS Free Tier for first 2 months

Anyscale

by Anyscale

Managed Ray platform for scaling AI training, batch processing and model serving with usage-based compute pricing.

Pay as you go; $100 starter credit

Apache Airflow

by Apache Software Foundation

Open-source platform to author, schedule and monitor data and ML workflows as Python code.

Free (open source)

Aporia

by Coralogix

AI guardrails and observability for LLM and ML systems. Now part of Coralogix (formerly Aporia).

AI usage from $1.50 per 1M tokens

Azure Machine Learning

by Microsoft

Enterprise Azure service for the end-to-end machine learning lifecycle, from data preparation to managed deployment and MLOps.

No extra charge for the service, pay for compute and Azure resources

BentoML

by BentoML

Inference platform for deploying and scaling any model, with autoscaling, observability and bring-your-own-cloud options.

Contact for pricing