Perpetual ML: Train, deploy and monitor models where your data lives
Perpetual ML is a machine learning studio that connects directly to your data warehouse for training, deploying and monitoring models.
What is Perpetual ML?
Perpetual ML is a web-based machine learning studio aimed at solo developers and data science teams. It covers the model lifecycle in one place: Auto Train builds models with the PerpetualBooster algorithm, experiment tracking records and compares runs, a model registry stores versioned models, and deployment serves them for batch or real-time inference. Monitoring watches metrics, data drift and model drift without requiring retraining or ground truth, and continual learning is described as reducing total training time from O(n^2) to O(n) across batches. Optimal business decisioning optimises a user-defined objective such as profit or risk. The product is built to run alongside existing data infrastructure rather than in a separate silo. It connects natively to Snowflake, so data stays in the warehouse under the same security and governance policies, with Databricks support listed as upcoming. Marimo notebooks are included for data exploration and model development. Pricing is pay-as-you-go: compute is billed at $0.00014 per vCPU second, persistent storage at $0.000069 per GB hour and egress at $0.10 per GB, while marketplace task pricing is free for 1-9 tasks, $9.9 per task per month for 10-99 tasks and $0.9 per task per month for 100+ tasks. AWS and Snowflake infrastructure costs are billed separately, and custom pricing is available on request.
Focus areas: Model Training.
Who is Perpetual ML for?
- Solo developers building and shipping ML models
- Data science teams that work primarily in Snowflake
- Teams that need to monitor model and data drift in production
How much does Perpetual ML cost?
Perpetual ML starts at $0.90/mo. Always check the vendor’s own pricing page before you buy.
| Plan | Price | What you get |
|---|---|---|
| 1-9 Tasks | Free | 1-9 tasks per month |
| 10-99 Tasks | $9.9 per task per month | 10-99 tasks per month |
| 100+ Tasks | $0.9 per task per month | 100+ tasks per month |
Is Perpetual ML worth it?
A warehouse-native ML studio that keeps training, deployment and monitoring close to Snowflake data, with usage-based pricing that starts free for small task volumes
Where Perpetual ML is strong
- Covers training, tracking, registry, monitoring and deployment in one interface
- Runs natively inside Snowflake so data does not leave the warehouse
- Usage-based pricing with a free tier for low task volumes
Where it falls short
- Databricks support is listed as upcoming rather than available
- AWS and Snowflake infrastructure costs are billed separately from the platform
How do you get started with Perpetual ML?
- 01Create an account on the Perpetual ML site
- 02Connect the platform to your Snowflake data warehouse
- 03Run an Auto Train experiment, then register and deploy the model
Perpetual ML: frequently asked questions
What does Perpetual ML include?
Auto Train, continual learning, optimal business decisioning, experiment tracking, a model registry, monitoring, deployment and Marimo notebooks, all from one web interface.
Does my data leave my data warehouse?
No. Perpetual ML is natively integrated with Snowflake, so data stays in the warehouse under the same security and governance policies.
How is pricing calculated?
Cloud usage is billed per vCPU second for compute, per GB hour for storage and per GB for egress. Marketplace task pricing is free for 1-9 tasks, $9.9 per task per month for 10-99 tasks and $0.9 per task per month for 100+ tasks.
Are infrastructure costs included?
No. AWS and Snowflake infrastructure costs are not included and are billed separately by AWS and Snowflake.
Perpetual ML alternatives
The closest tools in our directory, matched on category, focus areas and audience.
| Tool | Best for | Free plan | Starting price | Focus areas |
|---|---|---|---|---|
| Perpetual ML | Teams running ML workflows directly on Snowflake data | Yes | $0.90/mo | Model Training |
| Nebius | Cloud platform for building, tuning, and running AI models on NVIDIA GPUs. | Unknown | — | Model Training |
| Amazon SageMaker | AWS teams building, training and deploying ML models with governed data | Yes | — | Model Deployment, Model Training |
| Milk Infrastructure | Teams using the OpenAI SDK that want experiments from saved traffic | Unknown | — | Model Training |
| Imagga | Developers adding image tagging, search or moderation to apps | Yes | $79/mo | — |
| BentoML | Teams deploying and scaling model inference on their own infrastructure | Yes | — | — |
| Kubeflow | Kubernetes teams running ML training, pipelines and notebooks | Yes | — | — |
| vLLM | Self-hosting open-source LLMs with high throughput and low latency | Yes | — | — |
| NBULA | MENA enterprises and agencies needing Arabic-first, in-region AI deployment | No | — | — |
| Applied Compute | Enterprises building their own models | No | — | — |
| Clockwork | Operators of large GPU clusters | No | — | — |

