EN · 中文← all markets
On the public VRP benchmarks we ran, our engine went 9/9 on CVRPLIB-X and held the time-window deliveries Google's OR-Tools dropped (enterprise proof). Plan a route, lose the signal, it keeps navigating, the big maps need a connection, we don't. Fewer kilometres, fewer litres, fewer failed deliveries, on the maps you already use, at a fraction of the cost of the big names.
Same product worldwide. OpenStreetMap + live traffic. Benchmarks are measured on real graphs; route parity vs Google is live-tested in multiple cities.
On the public VRP benchmarks we ran, 9/9 on CVRPLIB-X against Google's OR-Tools, and we hold the time-window deliveries it drops (enterprise proof). Stop order stays within ~1.5% of best-known at 100 stops and ~0.9% at 200. Send your depot and stops; we optimise in your region before you integrate.
Plan a route, lose the connection, it keeps navigating, the big maps need a connection, we don't. The driver finishes the run on patchy mobile links or none.
MEASURED traffic workload: 302× faster per check; 97% skip when the jam misses your corridor, any city on OSM + TomTom.
MEASURED route update: 935 → 186 bytes (~80% smaller), less data on patchy mobile links everywhere.
Register vehicles globally, track live or GET /v1/assets/{id}/live on one API key. White-label it as your own.
Give us the depot and the day's stops; we return the shortest, fastest order to visit them, proven to match or beat the industry solver, in milliseconds.
A jam, a closure, a new pickup or a no-show becomes a small update, we re-plan only what changed, so the driver gets the new run quickly (compact delta on the wire where MEASURED).
The same distances and roads as the household-name maps, at a fraction of the per-call price, and your stops and locations never have to leave your systems.
OpenStreetMap roads, live TomTom traffic, Nominatim geocoding, use our map embed or yours. Hosted routing, no per-call lock-in; never locked to one map vendor.
REST over HTTPS. One key, free for 14 days, then pay. Running in minutes.
# Best order to visit the day's stops (depot + stops) POST /v1/optimize { "depot": [-26.20,28.04], "stops": [[...],[...],[...]] } # Plan one route · re-route when the plan changes (send only the change) POST /v1/route { "from": [...], "to": [...] } POST /v1/reroute { "from": [...], "to": [...], "prev_geometry": [...] } # Distance / time matrix for your own planning POST /v1/matrix { "depot": [...], "stops": [[...],[...]] }