We sell spatial decisions.
Not maps. Not tiles. Not lock-in.
Two pricing dimensions, transparently published. Metered API usage for context resolution, propagation simulation, and optimization runs — and a per-site calibration subscription for the loop that compounds in accuracy every month it runs on your facility.
Axis 1 · Metered usage
Per-call rates for /context, /simulate/*, and /optimize/* endpoints. Pay only for the spatial decisions you actually request.
Axis 2 · Site calibration
$200–$500/site/month for the calibration loop running on a specific facility. Models improve every day. The longer it runs, the more accurate it gets — for you alone.
Three ways to pay geog.ai
Most customers run all three: meter every sensor event through the API, subscribe a calibration loop to each operating site, and bundle the vertical-specific decision primitives that match their zooid deployment.
Spatial context resolution for any device. Returns a full SpatialState — terrain, jurisdiction, neighbors, wind vector, confidence score, full provenance.
GET /context per device, per eventGET /nearby, /within proximity queriesGET /jurisdiction at $0.0005/call (cached)The calibration loop running on a specific facility. Models improve every day. Switching cost increases every month — by design. This is the moat.
GET /model/calibration dashboard accessPhysics-based simulation jobs and optimization runs — plume, flood, RF coverage, noise contour, fire spread, sensor placement. Priced by model tier and complexity.
Per-endpoint pricing, in detail
Every published geog.ai endpoint with its billing dimension and rate. Async simulation jobs are billed on completion (delivered via webhook); failed runs are not billed.
| Endpoint | Type | Rate | Notes |
|---|---|---|---|
| Context & spatial query | |||
| GET /context | Per call | $0.001 | High volume, low margin — designed for every sensor event |
| GET /context/batch | Per device | $0.0008 | Discount per device when batched (≥10 device IDs per request) |
| GET /nearby | Per call | $0.001 | Returns proximity graph for the requested radius and type |
| GET /within | Per call | $0.001 | Spatial-containment query against an arbitrary geometry |
| GET /jurisdiction | Per call | $0.0005 | Cached aggressively; jurisdictions change rarely |
| GET /trajectory | Per call | $0.002 | MobilityDB-backed; includes terrain-adjusted route enrichment |
| Propagation simulation (async) | |||
| POST /simulate/plume | Per run | $0.05–$0.25 | Tier 1 (Gaussian, scipy) vs Tier 2 (AERMOD regulatory grade) |
| POST /simulate/flood | Per run | $0.10 | EPA SWMM watershed delineation + inundation mapping |
| POST /simulate/rf_coverage | Per run | $0.08 | Single-node coverage polygon at given RSSI threshold (ITM/Longley-Rice) |
| POST /simulate/noise | Per run | $0.08 | ISO 9613-2 octave-band noise contour |
| POST /simulate/fire | Per run | $0.15 | Fire-spread polygon + ember transport zone (terrain + fuel + weather) |
| RF mesh & coverage | |||
| GET /rf/coverage | Per call | $0.005 | Cached single-node coverage lookup at threshold |
| GET /rf/link | Per call | $0.002 | Point-to-point link budget between two nodes |
| POST /rf/mesh/viability | Per run | $0.15 | Full mesh connectivity matrix with failure-scenario probability |
| GET /rf/gaps | Per call | $0.01 | Coverage-gap heatmap across an area polygon |
| Optimization (highest margin) | |||
| POST /optimize/node_placement | Per run | $1.00–$5.00 | Recommends RF node placements. Consultants charge $50K+ for this manually. |
| POST /optimize/sensor_placement | Per run | $1.00–$5.00 | Vertical-aware sensor layout for a target coverage percentage |
| POST /rf/optimize/placement | Per run | $1.00–$5.00 | RF-specific placement with existing node constraints |
| Calibration (subscription) | |||
| Site calibration | Per site / mo | $200–$500 | Compounds in value — models improve every month they run |
| GET /model/calibration | Per call | Included | Dashboard reads are free for sites with an active subscription |
(vertical_id, site_id) calibration tuple.
$200–$500 per site, per month — priced by vertical complexity
A site calibration subscription is the right to have the geog.ai calibration loop running on one specific (vertical_id, site_id) tuple. The longer it runs, the more accurate your site-specific models get. Switching cost increases every month — by design.
Gas / plume sites
One propagation engine, one site, one regulatory regime. Typical for V01 gas-leak monitoring or V14 noise compliance at a fixed facility.
- One active engine (plume, RF, flood, or noise)
- Daily calibration cycle on receiver residuals
- ~90-day seasonal recalibration
- One vertical_id per site_id
- Audit log retained 7 years
Industrial / multi-hazard
Two-to-three propagation engines on one facility — typical for V31 oil & gas (plume + flood + RF) or V36 power generation (wind-load + RF).
- Up to 3 active propagation engines per site
- Daily calibration + 90-day seasonal cycle
- 1-year facility operating-mode learning
- Cross-engine residual sharing
- Priority queue for async simulations
Multi-vertical campus
All engines, all verticals at one site. Typical for V35 defense / tactical, V41 underground mining, or V50 wildfire regional command.
- All propagation engines (plume, RF, flood, noise, fire)
- Multiple
vertical_ids per site - Custom calibration cadence + named SRE contact
- Regulatory audit-trail export on demand
- SLA: 99.9% calibration freshness
Vertical-specific bundles for zooid deployments
Pre-packaged combinations of API endpoints + a calibration subscription, tuned to a specific zooid.net vertical. Cheaper than à la carte once you cross ~50,000 events / month, and shipped pre-wired into the matching ZOOID vertical.
Gas + Plume Bundle
V01 · Gas & Air QualityAdds geog.ai enrichment to every gas sensor event — plume forecast on alarm, jurisdiction-aware notification routing, upwind/downwind sensor cross-check.
RF + Coverage Bundle
V04 · Connectivity MeshMesh design and ongoing optimization API — coverage polygons, link viability, gap heatmaps, and quarterly placement-optimization runs included.
Flood + Watershed Bundle
V02 · Water & HydrologyHydrological context for every water sensor — watershed delineation, inundation forecast on rainfall threshold, and jurisdiction-aware overflow notifications.
Noise + Compliance Bundle
V14 · Noise & AcousticAutomated noise contour generation per source plus a continuous compliance check against the relevant jurisdiction's day/night limits.
Wildfire + Burn Management
V50 · WildfireFire-spread polygons, ember transport zones, and prescribed-burn containment geofencing. Ships with CAL FIRE/USFS jurisdiction overlays pre-loaded.
Custom Vertical Bundle
V05–V49 · On requestBundles for any of the other 50 ZOOID verticals — pick the endpoints you need, choose the calibration tier, and we'll quote a fixed monthly price.
Practical questions, answered up front
Things developers ask before signing up. If yours isn't here, email hello@geog.ai.
Free tier?
Yes. New API keys include 10,000 free /context calls, 200 free /jurisdiction calls, and 50 free Tier-1 simulations per month for evaluation. No credit card required to start. Site calibration subscriptions are paid only.
How are async simulations billed?
POST /simulate/* and POST /optimize/* return a job ID immediately. Billing happens only on successful job completion, when the result is delivered to your webhook or available via GET. Failed runs, timeouts, and validation errors are not billed.
What does “site” mean?
A site is one (vertical_id, site_id) tuple — typically one well pad, one factory, one substation, one watershed segment. A campus running multiple verticals (e.g. a refinery with both gas and noise calibration) is multiple sites.
Volume discounts?
Yes. Metered usage moves to tiered pricing automatically above 1M calls/month on a given endpoint group. Site calibration subscriptions discount 10% at 5+ active sites and 20% at 25+ active sites under one billing account.
Can I self-host?
Single-site enterprise deployments (air-gapped or VPC-isolated) are available for Tier-3 customers under a separate annual license. Pricing reflects the value the on-prem calibration loop generates — typically $50K–$250K/year per site. Contact us for a quote.
Currency, invoicing, payment?
All prices are in USD. Invoicing is monthly via Stripe with usage-based and subscription line items. Annual prepay is available with a 10% discount on subscription line items. ACH and wire are available above $5K/mo.
Two pricing dimensions. One spatial substrate.
Pay per call for spatial decisions. Subscribe per site for the calibration loop that compounds. Bundle vertical primitives when you cross the volume tipping point. That's it — no SKUs, no per-seat fees, no map quotas.