Available Models

Stratus-2 is the original ClearSKY Vision cloud-removal model and the ‘Product’ winner of the CASSINI Challenges. It leverages data from Sentinel-1 (ascending and descending), Sentinel-2, and Landsat 8/9 to produce cloudless, multi-spectral Sentinel-2 imagery. Thanks to its advanced data cleaning capabilities, Stratus-2 excels during periods with haze, thin cirrus clouds, and cloud shadows, delivering clear and accurate images. Its SAR (Synthetic Aperture Radar) capabilities make it particularly powerful in northern countries, where high SAR revisit frequencies significantly enhance image timeliness and provide more up-to-date imagery.

Stratus-2 (No Snow) is a specialized variant of the Stratus-2 cloud-removal model, designed for users who require consistent, cloud-free imagery without seasonal snow cover. Built on the same core architecture as Stratus-2, it leverages Sentinel-1 (ascending and descending), Sentinel-2, and Landsat 8/9 data to produce high-quality, multi-spectral images while reconstructing a snow-free surface, even when input data indicates snow presence. This ensures reliable and consistent imagery over time, making it particularly useful in regions with sporadic or transient snow, where intermittent snow cover can introduce inconsistencies in time-series analysis. It is especially recommended for agricultural applications, where fine-scale features such as crop health and soil conditions require stable, uninterrupted monitoring.

Nimbus represents the next generation in cloud removal technology and is currently in its final stages of testing. Engineered with modularity at its core, Nimbus integrates seamlessly with all available data sources on our platform, adapting to diverse input types for maximum versatility. By combining SAR and optical data, Nimbus achieves unprecedented accuracy in cloud removal, setting a new standard in the field. One of its most transformative advancements lies in how it processes SAR data compared to its predecessors, Stratus-2. With state-of-the-art SAR noise reduction algorithms, Nimbus extracts the maximum value from every image, even in challenging conditions.

Zephyr is a specialized model for harmonizing Landsat data to align with Sentinel-2 standards, enabling seamless integration of Landsat outputs into Sentinel-2 workflows. Designed for one-way harmonization, Zephyr ensures that Landsat data matches the spectral and spatial characteristics of Sentinel-2, providing consistent and comparable imagery for downstream analysis. By focusing on this targeted transformation, Zephyr unlocks the potential of integrating Landsat's longer temporal archive with the high-resolution capabilities of Sentinel-2, empowering users with harmonized multi-spectral data for diverse applications.