ModelMill Pricing
Four plans, one platform. ModelMill gives your team the tools to train Logic-Based Networks in the cloud and deploy them to the edge hardware already in your BOM, whether that's a CPU, an MCUs, or another 32-bit semiconductors. Choose the plan that fits your needs today, and scale when you are ready.
Test the advantages of LBNs.
One seat. Full access to every supported modality — sensor, tabular, time series, and audio. Upload your data, define your deployment target, and let ModelMill's automated training do the rest.
Demo plan includes:
- 1 user seat
- All supported data modalities
- Limited training runs
- Basic workspace
- Automated benchmarking
No model export. No deployment. This is your proving ground to see if LBNs work for your business case.
From experiment to edge deployment.
Everything in Demo, plus unlimited users, larger datasets up to 5 GB, and the feature that changes the conversation — C Code SDK export. Train your LBN, download the SDK, and deploy it directly to your edge device.
Everything in Demo, plus:
- Unlimited user seats
- Standard training runs
- 5 GB dataset capacity
- C Code SDK export
- Edge deployment
Get ideas (and colleagues) off the bench.
Everything in Pro, plus team workspaces — shared projects, shared visibility, shared progress — so that collaboration does not mean chaos. Optional SLA coverage for teams moving towards production.
Everything in Pro, plus:
- Team workspaces with shared project visibility
- Optional SLA
Your models. Your data. Your rules.
Everything in Pro Teams, plus prioritised training, guaranteed SLA, and custom capabilities — benchmarking visualisations, explainability modules, and export targets built around your requirements.
Everything in Pro Teams, plus:
- Prioritised training runs
- Guaranteed SLA
- Custom benchmarking visualisations
- Custom explainability modules
- Custom hardware export targets
Pricing built around your deployment
ModelMill pricing is structured around three elements: a base platform fee, training credits, and a deployment royalty. The right plan depends on your team size, training volume, and where your models need to run.
01
Platform fee
02
Training credits
03
Deployment royalty
FAQsFrequently asked questions
How much does ModelMill cost? 
ModelMill pricing has three components: a base platform fee for access to the training environment, training credits that cover compute during model training, and a deployment royalty applied when models are deployed to production hardware. Exact pricing depends on your plan tier and deployment scale. Book a demo to discuss your requirements.
Is there a free trial?
The Demo plan lets you explore the full platform — every supported data modality, automated training, and benchmarking — at no upfront cost. It is designed for evaluation, with limited training runs and dataset sizes. No model export or deployment is included in Demo.
How does the deployment royalty work?
When you deploy a trained LBN to production hardware — whether that is an edge device via the C Code SDK or a server via the inference API — a deployment royalty applies. This scales with your deployment volume. The royalty model means you pay in proportion to the value you extract, keeping upfront platform costs lower.
What hardware can I deploy LBN models to?
Any 32-bit processor or above. ARM (32- and 64-bit), RISC-V, x86, and ESP architectures are all supported. Models are exported as C code via the SDK, which integrates naturally with C++ codebases. No GPU, TPU, or specialist accelerator is required.
Can I upgrade my plan later?
Yes. Plans are designed to grow with your usage. Start with Demo to evaluate, move to Pro when you are ready to export and deploy, and scale to Pro Teams or Enterprise as your team and deployment footprint expand.
What data types does ModelMill support?
ModelMill trains LBNs across sensor data, tabular data, time series, and audio inputs. All supported modalities are available on every plan, including Demo.