Isaac Editor

Provides real-time inline suggestions for text completion, paraphrasing, and sentence refinement as users type in the editor. The system analyzes the current document context and generates suggestions via an integrated LLM (model unspecified), consuming daily AI tokens based on tier. Suggestions appear contextually without intrusive popups, allowing writers to accept or reject recommendations inline.

Enables users to upload academic papers and PDFs, then query them conversationally through an integrated AI chat interface. The system indexes uploaded documents and retrieves relevant passages in response to natural language questions, implementing a retrieval-augmented generation (RAG) pattern where the LLM generates answers grounded in document content. Supports file uploads up to storage tier limits (100 MB free, 1 GB basic, unlimited pro).

Provides built-in search functionality to discover academic papers and research articles directly within the Isaac Editor workspace. Users can search for relevant literature without leaving the editor, with results integrated into the document context. The underlying literature database source and search algorithm are undisclosed, but the feature aims to streamline literature review workflows by reducing context-switching between editor and external search engines.

Generates initial drafts of academic papers, essays, or sections based on user prompts and document context. This capability uses the integrated LLM to synthesize structured outlines or full draft text from minimal input, reducing the blank-page problem for academic writers. Available exclusively on the Pro tier, consuming unlimited AI tokens. The generation approach (prompt engineering, fine-tuning, or retrieval-augmented) is undisclosed.

Supports writing and AI assistance in 16 languages, with language-specific autocomplete, paraphrasing, and suggestions. The system detects document language and adapts suggestions accordingly, though the specific languages supported and language detection mechanism are undisclosed. Documentation states 'working on adding more' languages, indicating ongoing expansion.

Implements a daily token budget system that gates access to AI-powered features (autocomplete, paraphrasing, chat, generation) based on subscription tier. Free tier users receive 10 daily AI tokens, basic tier 50 daily tokens, and pro tier unlimited tokens. Token consumption per operation (e.g., one autocomplete suggestion, one chat message) is undisclosed, creating opacity around actual usage limits. Tokens reset daily, with no rollover or banking mechanism documented.

Provides cloud storage for documents with tier-based capacity limits (100 MB free, 1 GB basic, unlimited pro) and claims GDPR-compliant hosting on EU servers. Documents are stored remotely, enabling access from any browser without local installation. The storage architecture, encryption method (at-rest and in-transit), backup strategy, and data retention policies are undisclosed. No local-first or offline editing mode is documented.

Provides automated functions to streamline literature review workflows, including document organization, citation extraction, and synthesis. The feature set is explicitly incomplete ('More coming soon'), with specific automation capabilities undisclosed. This represents a planned capability rather than a fully implemented feature, indicating the product roadmap includes workflow orchestration but current implementation is minimal.