What Tools Does Mapillary Offer for Developers?

Mapillary gives developers three core building blocks: Mapillary JS for embedding street-level imagery in web map applications, image datasets for downloading and working with contributed street-level photos, and OpenSfM for structure-from-motion processing. These sit alongside hosting and GIS integration options listed on the site. The right entry point depends on whether you need to display imagery, analyze it, or process raw captures.

Mapillary JS: Embedding Street-Level Imagery

Mapillary JS is the developer library for building map applications that show street-level imagery. Use it when your application needs to let users view, navigate, or interact with street-level photos on a map — for example, a property or infrastructure inspection tool that lets users jump from a map location into the nearest street-level view.

The site lists Mapillary JS under its developer resources alongside the image dataset and OpenSfM, so it is positioned as the front-end viewing layer rather than an analysis tool.

Image Datasets

Mapillary states that its platform holds over 3 billion street-level images. Developers can access image and map data through the platform's data offerings. This is the option to consider when your task involves analysis rather than display — for example, training or evaluating a computer-vision model on street-level scenes, or studying road features across regions.

Two conditions matter here:

• Coverage varies by location. The site explicitly invites contributors to "fill gaps" by capturing imagery themselves, which signals that coverage is uneven. Check your target area before committing to a dataset-dependent workflow.
• Licensing terms apply. Mapillary describes its data as open data that supports decisions in urban and road construction, mobility and navigation, and autonomous driving. Review the actual data license before commercial use rather than assuming unrestricted reuse.

OpenSfM: Structure-from-Motion Processing

OpenSfM is Mapillary's open-source structure-from-motion pipeline, listed among its developer resources. Use it when you have your own image captures and need to reconstruct camera positions and 3D structure — the processing step that turns a sequence of photos into georeferenced, map-ready imagery.

A practical example: a team captures a road corridor with an action camera, then runs the images through OpenSfM to recover camera poses before contributing or analyzing the sequence. The site's "bring your own camera" positioning means the capture hardware is flexible; the processing pipeline is where OpenSfM fits.

Hosting and GIS Integration

Beyond the three core developer tools, the site lists street-level imagery hosting and ArcGIS integration. These matter if you want to store contributed imagery on the platform or surface it inside an existing GIS environment rather than building a custom viewer.

How to Choose

Start with the output you need. If users must see imagery, Mapillary JS is the direct path. If a model or analysis must consume imagery, look at dataset access and confirm coverage and license for your region. If you are producing imagery from your own captures, OpenSfM handles the reconstruction step. For teams already running ArcGIS, the integration option avoids building a separate viewer.

The site does not publish pricing or access-tier details for these developer tools on the pages referenced here, so confirm account and usage requirements through the developer resources and help center before planning a production build.