AI-powered QGIS plugin for automated map tracing and vectorization of geographic features from imagery.
AI Vectorizer is a Geospatial QGIS plugin from Bunting Labs that automates the tracing and digitization of raster maps (PDFs and GeoTIFFs) into vector line and polygon layers using a custom-trained machine learning model, with pricing starting at free via a trial tier. It is built for GIS technicians, cartographers, surveyors, and municipal planners who spend hours manually digitizing scanned maps.
Launched in January 2024 by San Francisco-based Bunting Labs, AI Vectorizer installs directly from the official QGIS plugin repository as "Bunting Labs AI Vectorizer." The workflow is designed around autocomplete: a user clicks two points to seed a line, and the AI extends the vector along the raster feature until it loses confidence, at which point the user can confirm, redirect with the shift key, or right-click to commit. According to Bunting Labs, the model keeps traced lines within the same pixel a human professional would choose on average, and can convert what used to be hundreds of clicks into a two-click, sub-minute operation—roughly a 10x throughput improvement over non-ML predecessors like AutoTrace (last updated 2015) and ContourTrace (2011).
Based on our analysis of 870+ AI tools, AI Vectorizer is one of the few GeoAI products purpose-built for the QGIS ecosystem rather than a general computer-vision tool retrofitted to maps. Inference runs remotely on Bunting Labs' encrypted servers, so users do not need high-end GPUs locally—if a machine can run QGIS, it can run the plugin. It sits alongside the company's AI Georeferencer (which currently aligns scanned maps to within 5 meters, targeting sub-1-meter accuracy) and the Kue AI and Mundi web GIS products. Compared to manual digitizing workflows or legacy path-finding plugins, AI Vectorizer is the more modern choice for teams processing large backlogs of scanned parcel maps, utility as-builts, contour sheets, or historical cartography.
Was this helpful?
Users click once at the start of a raster line and once a short distance along it to give the AI directional context. The model then traces as much of the feature as it can confidently follow and pauses for user confirmation, turning hundreds of manual vertex clicks into a two-click operation.
Activating the plugin on a polygon layer lets the AI walk the border of a closed feature and fill the interior as a single geometry. This eliminates the tedious vertex-by-vertex capture of parcel boundaries, building footprints, or landcover polygons.
When autocomplete gets distracted by a nearby line or drifts off-center, holding shift lets users click the last good vertex to truncate the trace, then click past the difficult region. Releasing shift re-engages autocomplete from the new position, so corrections happen in-flow without restarting the feature.
The trained model runs on Bunting Labs' web servers rather than the user's machine, with encrypted traffic between plugin and server. This removes any local GPU or memory requirement—if QGIS runs, the plugin runs—and centralizes model updates so users always get the latest weights.
Bunting Labs trained a proprietary architecture specifically to isolate raster lines on diverse map styles, prioritizing precision over recall so that only confident segments are emitted. On average the model stays within the same pixel a GIS professional would choose, though the company notes it has not yet exceeded human performance.
$0
$29
Custom (contact Bunting Labs)
Ready to get started with AI Vectorizer?
View Pricing Options →We believe in transparent reviews. Here's what AI Vectorizer doesn't handle well:
Weekly insights on the latest AI tools, features, and trends delivered to your inbox.
As of early 2026, Bunting Labs continues to actively develop AI Vectorizer with iterative model improvements pushed via their cloud inference pipeline—users automatically receive updated weights without reinstalling the plugin. Throughout 2025, the company expanded its GeoAI product suite alongside AI Vectorizer, including updates to AI Georeferencer (targeting sub-1-meter alignment accuracy) and the Kue AI and Mundi web GIS platforms. The QGIS plugin remains available through the official plugin repository with ongoing compatibility updates for recent QGIS releases. No major pricing restructure has been publicly announced since launch; the credit-based freemium model introduced in January 2024 remains in effect. Users should check buntinglabs.com/blog and the QGIS plugin changelog for the latest feature additions and model accuracy benchmarks.
No reviews yet. Be the first to share your experience!
Get started with AI Vectorizer and see if it's the right fit for your needs.
Get Started →Take our 60-second quiz to get personalized tool recommendations
Find Your Perfect AI Stack →Explore 20 ready-to-deploy AI agent templates for sales, support, dev, research, and operations.
Browse Agent Templates →