Frequently Asked Questions

You can access cloud-free Sentinel-2 images via our cloudless data fusion service on order.clearsky.vision. Simply sign up on our platform to gain access or request sample data through our online demo form.

Our service uses advanced cloud removal and data fusion techniques by combining data from multiple satellites (including cloud penetrating SAR imagery) and deep learning models to produce clear, cloud-free imagery. The images are, therefore, cloud-free since we clean up optical imagery to the extent it's possible while fusing cloud penetrating SAR imagery into the optical imagery.

Our service leverages data from several satellites, including Sentinel-1, Sentinel-2, Sentinel-3, Landsat 8/9, CBERS4/4A, and Amazônia-1, ensuring comprehensive coverage and up-to-date imagery. We are consistenly adding new satellites into the mix.

When generating cloudless data for a specific order, we allow users to select additional satellites beyond a set of required ones. These required satellites form the foundation for ensuring high-quality images. However, based on the specific geometries of the order, users can add certain additional satellites to further improve image quality and timeliness. For example, in most regions of Brazil, satellites such as CBERS-4 (AWFI & MUX), CBERS-4a (WFI & MUX), and Amazônia-1 can be selected as additional input satellites. Once added, we will use all available imagery from these satellites within the order's timespan to meet the requested image frequency. Regardless of whether the frequency is every second day or every seventh day, all available data will be utilized to optimize the results.

The ‘Image Frequency’ refers to the number of days between the creation of images for a specific order. It can be adjusted to any value between 1 and 7, where 1 represents daily image production, and 7 signifies weekly image production. If you need another frequency of images, contact support to hear more about other options.

When placing an order with us, users can specify how long they need the ordered images to be stored. By default, every image is automatically stored for at least one month from the production date. If longer storage is required, users can extend this period by selecting the number of additional months needed. This means that images from an order will be deleted at different times but storage months can be used to delay this deletion.

We have separated download requests from image storage. While images are stored for at least one month after production (plus any additional storage months selected), the download credits—measured in 'km²'—do not expire for a full year. This means that if your images are produced and then deleted before you've downloaded all of them, the unused credits remain available. You can use these credits to reacquire the images by placing a new order within the same year.

Most orders are processed within 5 hours, but large, complex requests may take up to 30 hours. Contact us if normal processing speeds are a concern.

Customers may cancel an order while its status is “Review” as production has not yet started. Once an order is confirmed, cancellation is no longer possible due to incurred processing costs.

Yes, but you must cancel your recurring order before the next billing cycle starts.

No, Tile orders are fixed in size (5120×5120 pixels), and partial tile orders are not possible.

The minimum order cost for our satellite imagery services is 50 EUR. To maintain an active account, a cumulative monthly spend of at least 50 EUR is required across all orders. If the total spend falls below this threshold, a flat charge of 50 EUR will be applied to your account.

Our pricing model includes a minimum cost of 50 EUR for any new tasking order during the first month. If your order’s calculated cost is below this amount, the minimum charge ensures a baseline for processing. For recurring orders in subsequent months, this minimum does not apply - you’ll only be charged based on actual usage or tasking requirements.

Each cloud-removal model has a base price, which includes the cost of required satellite inputs. For models that support optional satellites, their prices are added to the base price for each optional satellite used.

For example: Stratus-2 has a base price of 0.20 EUR/km² (including required satellites).

Our tiered discount system, now based on Usage Units (UU), calculates monthly discounts from composite and tile orders. Orders contribute to cumulative monthly totals, unlocking higher discount tiers. Discount Tiers range from 0% (0–5,000 UU) to 98% (500,000,001+ UU). Stratus-2 composite orders earn 1 UU/km²; Tile orders earn 0.1 UU/km². Usage units reset monthly, and all orders contribute toward your discount for that month.

No, historical data is not inherently cheaper than live or recurring data. However, it is easier to obtain larger bulk discounts on historical data. When users select multiple historical months, these are bundled into a tiered discount structure, meaning that large orders receive progressively higher discounts. In contrast, live/recurring data involves more complex processes, such as continuous real-time data acquisition, frequent data loading, and processing. While the cost per unit of historical data is identical to live data, the ability to order in bulk and benefit from tiered pricing makes historical data more affordable for large-scale requests.

Overlapping composite orders with identical satellites and revisit frequency will be deduplicated (not billed twice). However, Tile orders are always billed separately, even if overlapping.

API requests require an API key. You can obtain one by signing up and generating a key from order.clearsky.vision.

Default rate limits: 4 concurrent requests and 4 requests per second. Higher limits may be available upon request.

Live Data is available by 10:00 AM CET the next day after acquisition. Historical Data is processed upon order placement and delivery time depends on AOI size.

No, all cloudless data has been produced in the same way. This is to ensure consistency throughout our service. However, this also means historical data is produced without any future insights and it’s only backward-looking. It will not interpolate between the date and what we know it will become in the future because we will not ingest future data even if it is available.