Pricing for catalog collections depends on the provider. For example, low resolution images cost less than high resolution ones.
Specify order parameters in the console to see the final price.
Pricing for tasking collections depends on the requirements you specified for your order. For example, a mono acquisition mode will cost less than stereo or tristereo modes. Different providers have different parameters for you to specify.
Submit a tasking order request in the console to receive a quotation.
Pricing for data and processing blocks is displayed on their corresponding pages on the marketplace. It can be set for different units, but most often it's set for km2.
Configure your job to see an estimation calculated from the following costs:
How much data will be consumed through a data block. Pricing for data blocks depends on technical specifications, such as spatial, spectral, and temporal resolution.
How much data will be processed through a processing block. Pricing for processing blocks depends on algorithm complexity, such as band math, deep learning, and interferometry.
How much machine time and what machine type will be used during data consumption and computation.
Note that infrastructure costs aren't displayed in the job estimation, but they're usually negligible. You can check them in the job dashboard after running the job.
All marketplace blocks have a machine type assigned to them by their provider, depending on the algorithm complexity. The chosen machine type affects the infrastructure costs for running a job with this block.
|Machine type||CPU||GPU||RAM||Price per hour|
|0.5||None||2 GB||4 credits|
|1||None||5 GB||9 credits|
|2||None||10 GB||19 credits|
|4||None||20 GB||38 credits|
|8||None||40 GB||76 credits|
|16||None||80 GB||152 credits|
|4||1||20 GB||99 credits|
|8||1||40 GB||258 credits|
You can choose a machine type for your custom block. Different use cases will require different machine types:
- Use with a data block that downloads images.
- Use with a processing block with a very basic algorithm.
- Use with a processing block with a basic algorithm. For example, it's used with NDVI.
- Use with multi-band imagery analysis that can't be parallelized with raster tiling.
- Use with advanced algorithms. For example, it's used with K-means clustering.
- Use with GPU intensive ML algorithms. For example, it's used with Super-resolution Sentinel-2.
All CPUs and GPUs are charged a minimum of 1 minute. For example, if you run your machine instance for 30 seconds or less, you'll be billed for 1 minute of usage. After 1 minute, instances are charged in 1-second increments.
Any fractional credit usage gets rounded up to 1 credit. For example, 0.1 credit is billed as 1 credit.