Lexor Documentation Manual

1. Overview

Lexor is a desktop-based, AI-powered writing application designed for writers of all experience levels. It is sold as a one-time purchase; however, AI functionality requires the user to provide their own API key. AI usage is billed separately on a pay-per-use basis by the selected API provider.

Lexor’s primary goal is to make writing faster, easier, and more enjoyable by removing friction such as writer’s block, manual rewriting, and context management. It supports a wide range of writing formats, including fiction, lyrics, structured emails, and general prose.

2. AI Usage & API Fundamentals

Important: API Key Requirement

Lexor does not bundle AI usage. Users must supply an API key from a supported provider in order to generate, edit, or analyze text.

Supported Providers

  • OpenAI – Offers multiple cost-efficient and high-quality language models.
  • xAI – Provides fewer content restrictions, suitable for uncensored or darker material.

Users can switch providers at any time in the settings.

Cost Control & Transparency

Lexor tracks token usage in real time and displays estimated costs directly in the interface. This ensures users always understand how much they are spending.

Budget controls include:

  • Per-project spending limits (e.g., $5 maximum)
  • Warnings when approaching limits
  • Automatic prevention of accidental overspending

3. Core Architecture & How Lexor Thinks

Lexor’s AI operates using three layers of context, ordered by importance:

  1. Selected Text – Highest priority; directly edited or expanded.
  2. Active Canvas Content – Surrounding text that informs tone and continuity.
  3. Context Library Files – Optional background knowledge (characters, lore, rules).

Only files explicitly enabled or triggered are sent to the AI, minimizing token usage and maximizing accuracy.

4. Main Components

4.1 Writing Canvas (Primary Workspace)

The Writing Canvas is Lexor’s core environment. It behaves like an infinite page that grows as you write, prioritizing content creation over visual formatting.

Formatting is previewed using lightweight symbols rather than full styling, keeping focus on writing flow instead of layout.

Core Canvas Tools

  • Generate: Continues writing in the existing style and tone. Supports bridging, allowing users to place the cursor between two text sections and have the AI seamlessly connect them.
  • Guide Canvas: Defines a global system instruction that applies to all Canvas generations (e.g., writing style, tone, genre rules).
  • Create (requires selected text): Rewrites text using transformation presets such as:
    • Humanize (removes generic AI phrasing)
    • Symphonize (converts text to lyrics)
    • Humorize
    • Summarize
    (Presets are fully customizable.)
  • Edit (requires selected text): Applies structural improvements such as:
    • Expand
    • Concise
    • Fix All (grammar, punctuation, capitalization)
  • Alternate (requires selected text): Generates multiple alternative phrasings for the same idea. Results can be regenerated indefinitely.
  • Manifest (requires selected text): Expands shorthand ideas into polished prose, turning rough concepts into complete narrative passages.
  • Prompt (requires selected text): Allows direct custom instructions. The AI performs the edit silently without extra commentary.
  • Chapter Generator: Accepts bullet-point outlines and expands them into fully written, connected chapters.
  • Translate: Opens a translation workspace for converting Canvas content into any language and exporting localized files.
  • Copy / Export: Copy content to clipboard or export to common text formats.
  • Export to Context: Saves the current Canvas content as a file inside the Context Library.

4.2 Memory System (Token Optimization)

The Memory system summarizes key story events into compact references. These summaries are used instead of full text when sending context to the AI.

  • Adjustable summary length (e.g., 3–5 paragraphs)
  • Can be auto-generated or manually edited
  • Reduces token usage while preserving narrative continuity

4.3 Context Library (Knowledge Management)

The Context Library functions as an organized file system for background material such as Characters, Lore, Locations, and Rules.

Files can be toggled on or off for AI visibility.

Keyword Activation

Files can be assigned keywords. When a keyword appears within the AI’s active context window, the associated file is automatically included, ensuring accurate references without manual intervention.

Import & Auto-Organization

Users can import zip folders of existing content. Lexor’s AI analyzes and restructures them into organized files.

4.4 AI Chat

AI Chat provides a standalone conversational interface separate from the Canvas.

  • Create custom AI roles (e.g., Reviewer, Editor, Worldbuilding Consultant)
  • Discuss Canvas or Context Library content
  • Save, reset, or manage chat history

This mode is ideal for brainstorming, critique, and planning.

4.5 Roleplay Layout

Lexor offers two interface layouts: Writer Layout (default) and Roleplay Layout.

The Roleplay Layout allows users to interact directly with story characters. The AI can fully embody characters, enabling dialogue testing, character consistency checks, and exploratory roleplay.

5. Customization & Quality of Life

  • Font type and size adjustments
  • Fullscreen distraction-free mode
  • Project UI preset importing (without importing content)

6. Settings & Configuration

Settings Include:

  • System prompt management
  • API provider selection
  • Budget and usage tracking
  • Project preset importing