Skip to content
Aquedra Run

Pricing

GIS tools for agents, with the price up front.

You pay per call from prepaid credits (1 credit = US$1). Every call has its price before it runs, and you are charged only for what works.

Price by tool
ToolWhat it doesREST routeMCP toolPrice
Coordinate conversionConverts points between decimal degrees, degrees-minutes-seconds and UTM (WGS84).POST /v1/geo/coordinatesconvert_coordinatesUS$0.001per unit: up to 1,000 points
GIS measurementMeasures the geodesic area, perimeter and length of each feature of a GeoJSON, with totals.POST /v1/geo/measuremeasure_geojsonUS$0.001per unit: up to 1,000 geometries and 10,000 vertices
GIS reprojectionReprojects a GeoJSON between coordinate reference systems such as EPSG:4326, EPSG:3857 and the WGS84 UTM zones.POST /v1/geo/reprojectreproject_geojsonUS$0.001per unit: up to 1,000 geometries and 10,000 vertices
GIS transformationApplies buffer, simplify, centroid, bounding box, convex hull or clean to a GeoJSON.POST /v1/geo/transformtransform_geojsonUS$0.001per unit: up to 1,000 geometries and 10,000 vertices
GIS file conversionConverts a CSV, KML, GPX, GeoJSON or zipped Shapefile to GeoJSON, KML or CSV.POST /v1/geo/convertconvert_geo_fileUS$0.01per file, whatever its number of features
Extended timeEvery call includes 3 s of compute. With max_seconds from 4 to 30 (extended time) each extra second costs US$0.0005: it is held at the start and what you do not use is refunded. If the time runs out, the price and the whole held time are charged; if memory runs out or the call is cancelled while running, the price and the time used are charged.US$0.0005

You pay only for what works

With the default 3 s a call is charged only when it succeeds, and invalid input is never charged. With dry_run you see the price without running or charging anything, and with max_cost_usd you refuse any call that would cost more than you set.

Inference (own GPUs, best effort)

Inference prices per million tokens
ModelUseInput (per million tokens)Output (per million tokens)
qwen3-8bGeneral chat and reasoning; the default model.US$0.18US$0.70
qwen2.5-3bpricing.inference.models.qwen2.5-3bUS$0.05US$0.10
qwen2.5-coder-7bpricing.inference.models.qwen2.5-coder-7bUS$0.025US$0.05
qwen2.5-vl-7bpricing.inference.models.qwen2.5-vl-7bUS$0.18US$0.69

Minimum per call: US$0.0005. The maximum cost is held before the call runs and what you do not use is refunded.

If the GPU is busy, the answer is 503 with Retry-After and nothing is charged: retry after those seconds. We do not store your prompts or the answers.

Coming soon: isolated Python, document conversion, browser, and terrain analysis.

Start with an account

Create an account and connect your agent over API or MCP. Credits are not on sale yet.

Create an account

For your agents