May 13, 2025
Last updated
Last updated
Learn about the enhancements, new features, and resolved issues as part of this release.
We’ve officially transitioned to a new, production-ready domain: . This change marks a significant step toward platform stability and scalability.
Why this matters: The new domain provides a cleaner, more reliable access point to the RapidCanvas platform.
Transition period: Users can continue accessing the platform via both and . However, moving forward, will be the primary domain for all users.
This update ensures a consistent and robust experience across sessions.
Introducing RapidRAG libraries — a specialized set of helper functions developed to improve the performance and flexibility of Retrieval-Augmented Generation (RAG) workflows on RapidCanvas.
Environment-Level Configuration
A new checkbox, "rapid-rag libraries", is now available in environment settings.
When selected, the libraries are included for all recipes in the project.
Ideal for consistent usage across multiple RAG recipes.
Recipe-Level Configuration with Overrides
Enable RapidRAG libraries at the individual recipe level under the Requirements section.
Ensures the recipe runs in an isolated, temporary environment, preventing version conflicts.
Any libraries specified here will override those defined at the environment level.
We've improved the Requirements UI in code recipes to simplify dependency management:
Two new tabs:
Environment Requirements
View all Python libraries installed at the environment level.
Contains the environment-level RapidRAG checkbox.
Libraries here are read-only from this tab.
Recipe Requirements
Define Python libraries specific to a single recipe.
Includes a separate RapidRAG checkbox for isolated usage.
If left empty, the system defaults to the environment-level settings.
Users will now see an informational alert when an environment is being overutilized—either by their own task or by other processes within the project. This proactive notification helps users better manage resources and switch to more available environments when needed.
Disk space can now be modified after environment creation, providing greater adaptability for changing storage needs.
Default disk space has been increased from 100 GB to 200 GB.(at deployment level)
Default memory has been upgraded from 48 GB to 56 GB.(at deployment level) These changes ensure smoother execution of data-intensive tasks.
A new section in the Environment Configs side sheet now displays current disk usage, making it easier to monitor and manage available storage.
Users can now easily insert Text Creation blocks in the canvas using a sample syntax snippet, available via the Syntax button in the code recipe workspace.
The file size of uploaded text files is now shown in the side sheet—just like with datasets—enhancing visibility and control over large text inputs.
The Prediction Log Details window has been redesigned for better usability:
Request, Response, and Logs sections can now be expanded or collapsed individually.
Scrollbars have been added to each section and to the overall window for easier navigation through long outputs.
A Clear All option has been added to the Freeze Columns dropdown in the Rapid model recipe, allowing users to quickly deselect all frozen columns in one click.
Users can now view the "Last Updated On" timestamp for files in the Artifacts folder, aiding better tracking of file modifications and versioning.
The embedded Streamlit app has been updated from version 1.14.0 to the latest available release, offering enhanced performance, new features, and improved compatibility.