AI Central
AI Central
Breadcrumbs

AI Central V4.2

Release Date: June 2026
Version: AI Central 4.2
Version Type: Feature Upgrade / Experience Optimization / Stability Enhancement

Overview

AI Central 4.2 focuses on enterprise-scale AI adoption, continuously enhancing five key areas: intelligent agent execution, knowledge services, data integration, workflow orchestration, and platform governance. This release strengthens both capability depth and platform controllability, helping enterprises build, deploy, and operate AI applications more efficiently.

This version includes 40+ new features and optimizations. Among them, Super Agent, Pipeline, AI Slides, and AI Translator represent the most impactful upgrades: Super Agent improves end-to-end task execution capability, Pipeline enables knowledge engineering workflows, while AI Slides and AI Translator significantly accelerate reporting and cross-language collaboration, forming a complete productivity loop from “task execution” to “content delivery”.

Key Highlights

1. Super Agent Upgraded to an Executable Intelligent Agent

  • Supports multi-step task execution workflows, enabling continuous reasoning and action in complex tasks.

  • Supports task recognition and multi-agent collaboration, improving efficiency in complex problem-solving.

  • Newly added capabilities: web search, chart generation, file parsing, and file generation.

  • Supports voice input, model switching, and conversation history management.

2. Pipeline Enhanced for Knowledge Engineering

  • Introduced Preprocess Pipeline for document preprocessing, parsing, and structuring.

  • Introduced Retrieval Pipeline supporting hybrid vector search and full-text search.

  • Enhanced reranking and context construction capabilities to improve answer accuracy and relevance.

  • Added retrieval testing and observability features for continuous optimization and evaluation.

  • Designed for enterprise multi-source document ingestion scenarios, significantly shortening the time from “data ingestion” to “usable knowledge”.

3. Standardized Knowledge Service Exposure

  • Supports Dify External Knowledge API standard integration.

  • Supports OpenAPI-based knowledge retrieval services for easy integration with external systems.

4. Enhanced Data Integration and Modeling Capabilities

  • Newly supports Snowflake, Fabric, and PostgreSQL data sources.

  • Supports schema-level configuration for SQL Server.

  • Introduced unified data catalog management to improve asset organization and discoverability.

5. Enhanced Workflow Orchestration Capabilities

  • Introduced workflow-level environment variables and session variables.

  • Supports dynamic configuration of knowledge retrieval parameters (e.g., TopK, thresholds).

  • Added Python Sandbox secure execution environment.

  • Added Auto Organize node for automatic layout optimization, improving orchestration efficiency.

6. AI Slides and AI Application Expansion

  • Newly added or enhanced AI Slides, AI Translator, AI Reading, and Deep Research capabilities.

  • AI Slides can generate presentations based on enterprise knowledge assets, reducing report preparation time from hours to minutes.

  • AI Translator supports multi-format document and scanned document processing, improving cross-language collaboration efficiency and consistency.

  • Unified AI application entry and resource management for better application access and operation efficiency.

7. Enhanced Agent and Ecosystem Capabilities

  • Added Agent sharing links for quick distribution and access.

  • Supports MCP Server publishing and invocation, improving external capability integration flexibility.

  • Supports unified mounting of multiple tools, skills, and data sources for flexible scenario orchestration.

8. Expanded Model Support

  • Added support for Claude and Google model families.

  • Supports multiple embedding models for vector representations.

Feature Enhancements and UX Improvements

Knowledge Base Experience Optimization

  • Improved interaction flow for knowledge base and file selection.

  • Supports folder-level batch selection and more flexible selection strategies.

  • Supports direct navigation from knowledge base to translation capabilities, improving knowledge processing efficiency.

Deep Research Experience Optimization

  • Improved task decomposition and execution logic.

  • Supports editable sub-questions and asynchronous execution.

  • Enhanced stability and controllability of complex research tasks.

AI Translator Enhancements

  • Improved multi-format translation for DOCX, PPTX, XLSX, and PDF (including scanned documents).

  • Optimized terminology mapping logic for improved translation consistency.

  • Supports batch translation with unified output for better global collaboration.

Platform Governance and Observability Enhancements

  • Enhanced operation logs and invocation logs.

  • Added token usage statistics and analytics capabilities.

  • Improved RBAC and asset authorization for better enterprise governance.

Stability and Bug Fixes

  • Fixed skill list display issues in some orchestration flows.

  • Fixed agent invocation anomalies in complex scenarios.

  • Fixed edge cases in knowledge retrieval and document processing.

  • Improved system stability and response performance under high concurrency.

  • Other known issues resolved.

Business Value

  • Faster delivery: Shorter path from capability building to application deployment, reducing implementation complexity.

  • Higher quality: Improved stability in knowledge retrieval, content generation, and workflow execution.

  • Easier operations: Unified entry points, resource-based management, and enhanced operational analytics.

  • Stronger governance: Improved permissions, auditing, and usage tracking for enterprise compliance.

Upgrade Recommendations

  • Enterprise administrators are advised to verify model availability scope and role-based access control after upgrading.

  • It is recommended to enrich terminology glossaries and prompt templates for high-frequency scenarios to ensure more stable outputs.

  • Use logs and token analytics to continuously optimize workflow and retrieval parameters.

Notes

This document is a public-facing release note intended for official website updates, customer announcements, pre-sales materials, and marketing communications. For project delivery documentation, additional details such as deployment architecture, compatibility matrices, and implementation boundaries should be added based on this version.