The main purpose of this release is to make AI-powered conversions more stable and predictable for end users, and to give template editors additional visual and structural control inside the editor.

We introduced internal fallback behavior for the LLM models used in the first conversion step, so users are less likely to encounter failed or incomplete AI conversions. We also extended the editor’s flexibility with conditional background colors and new section-level settings in the AI Editor, helping users build clearer, more readable templates.

More reliable AI conversions

How it works
When the LLM cannot complete the first step of an AI-based conversion, Ascendix Composer now uses an internal fallback path to keep the conversion going rather than failing. From the user’s perspective, this makes AI conversions more likely to complete successfully without additional manual retries.

Why
To reduce failed AI conversions and provide a more stable experience when using AI-driven features, especially for users who rely on AI to prepare content and layouts.

Conditional background color in templates

How it works
You can now configure background colors to change automatically based on conditions (for example, specific values or states in your data). Once set, the background color updates according to the configured rules at the time of generation.

Why
To help users visually highlight important data or states (such as status, thresholds, or categories) without manual styling, and to make generated reports easier to scan and interpret.

New “Section settings” in the AI Editor

How it works
The AI Editor now provides dedicated section-level settings. For each section, users can adjust configuration options (such as structure and behavior) directly in the editor rather than relying solely on global settings.

Why
To give users finer control over how individual sections are organized and rendered, and to simplify tuning AI-generated layouts without needing workarounds or template-level rework.