Flexible pricing and four ways to run Pointly.

No. A short catalog that only splits ground from vegetation, a standard classifier with the classes a power line survey needs, or a long custom catalog: the rate per million points is the same. The only thing that changes it is whether the classifier is a standard one or trained on your own catalog.

Everything comes back to one rate per million points: 0.15 € on a standard classifier, 0.20 € on one trained on your own catalog. For a managed project we translate that into your unit, per km² or per running kilometer, after looking at your sample data. One unit, fixed price before the work starts.

Points, counted once on what you upload. The output comes back free of any further count, and it is yours to keep and to move at any size.

The upload is blocked at the limit and nothing is billed on top, so the month lands on the package price you agreed. Volume is checked before every upload, so nothing is reconciled afterwards.

Both. A fixed budget for a year, drawn down over the period, or another package added alongside the one you have.

Yes, on the price per km², and it grows with the size of the project. Put your area and density into the calculator above and it shows the figure.

When a standard classifier does not carry the classes you work to. M+ is Professional M's volume with one classifier trained on your own class catalog and post processing tuned to your data, so every upload after that runs against your classes rather than ours.

Training that classifier is quoted per project and is not part of the subscription. The monthly price covers running the trained model. What it costs to build one depends on how far your classes sit from the existing training base, and we name that figure after looking at a sample of your data.

The same route carries more than a retrained model. We set up a workflow around it — an existing model optimised for your sensor and density, or automatic vectorization that returns object instances and models rather than classified points. Both are scoped per project. Tell us your class catalog.

The packages are sold to companies, and you upgrade to one from inside the platform once you have an account — a free one is enough to start with.

Yes. When a class you need is not in a standard model we train one that has it. Adapting an existing classifier to a new catalog is inside a single day, because the training base is modular. Training a new model is a few weeks, and the time goes into generating the training data. Quoted per project, then run at 0.20 € per million points. Tell us your class catalog.

Yes. Anything a classified cloud describes can come back as geometry, and the toolset is not a fixed catalog: tower positions, conductor runs, insulator suspension points and road assets are simply the ones asked for most, and an object nobody has asked us for yet is a scoping question. Out as DXF, Shapefile, CSV or GeoJSON, the BIM-ready vectors a model is built from rather than the finished BIM file or a mesh. Quoted per project, because it is a custom implementation today. It is not a list price item. Tell us what you need extracted.

Your original LAS or LAZ file, untouched, with the classification added as a dimension. Full resolution, nothing downsampled and nothing compressed. Where vectorization is part of the job, the objects come as separate files.

Yes. The free account runs the standard classifiers on your own data, three files at 15 MPoints each, and we raise the limit on request.

On Microsoft Azure inside the EU by default. Where a project needs a different region we set up a separate Pointly tenant with storage and processing at any available Azure location, which already runs in several regions including the US.

Yes, with the Pointly Box. It runs the full stack inside your network, air gapped if required, and the license sits on a hardware key so nothing has to reach us.