QNAP has released its core search application, Qsirch 7.1.0, which significantly strengthens semantic search capabilities through advanced AI integration. This update incorporates on-premise Vision Language Models (VLM) and Large Language Models (LLM), elevating data management within Network Attached Storage (NAS) to an enterprise-grade level.

AI Integration and the General Release of Premium Features

AI Mode Revolutionizes the Search Experience

On August 12, 2026, QNAP Systems, a major innovator in computing and storage based in Taiwan, announced the launch of Qsirch AI Mode as a major upgrade to its core search tool. Traditional search methods relied heavily on exact keyword matching or manually assigned tags, but AI Mode utilizes natural language processing to understand the actual meaning of content. Users can now find files by describing them in plain language, such as searching for “sunset photos at a beach from last year” or finding specific audio segments where a particular project was mentioned. This semantic understanding applies across documents, images, and videos, allowing the AI to index data based on context rather than just filenames. Refer to the following diagram for the workflow of AI-driven summarization and search. .

Strategic Expansion Through Free Premium Access

In a move to accelerate the adoption of AI-driven knowledge management, QNAP implemented a bold strategy by making all Qsirch premium features free of charge. Effective October 1, 2025, the previous premium licensing program was abolished, allowing all users—including those formerly on the Qsirch Lite tier—to access advanced search and AI filtering functions without additional costs. This transition is part of QNAP’s vision to evolve the NAS from a simple storage device into a comprehensive AI Knowledge Management Platform. Furthermore, for users on older operating systems in Japan and worldwide, QNAP released Qsirch 5.3.3 in late August 2025 to lift licensing restrictions and provide a modern search experience across its broader product lineup.

Advanced Semantic Understanding via VLM and LLM

Multimodal Search Across Four Major Media Formats

The powerful capabilities of Qsirch 7.1.0 are driven by the integration of multimodal AI architectures. Specifically, Large Language Models (LLM) are used to analyze the context of documents such as PDFs, DOCX, and PPTX files, identifying relevant paragraphs based on meaning rather than just character strings. Simultaneously, Vision Language Models (VLM) visually analyze image content, eliminating the need for manual tagging or album organization. This integration extends to video (MP4, MOV) and audio (WAV, MP3, M4A) files, which the AI understands and indexes at a semantic level. This cross-media search capability allows professional users in Japan and globally to dramatically reduce the time spent navigating vast digital archives.

Enhancing Efficiency with Timeline Sync and Transcription

Qsirch 7.1.0 introduces highly practical tools for managing audiovisual data: synchronized timelines and automatic transcriptions. The AI automatically generates text transcripts for video and audio files, allowing users to jump directly to a specific second in the media just by clicking a piece of the transcribed text. This is particularly useful for finding specific topics within long seminar recordings or meeting logs. Additionally, the current line being spoken is automatically highlighted during playback, making content verification effortless. By treating media assets as searchable text data, QNAP provides a massive benefit to creative professionals and legal or academic researchers who deal with high volumes of recorded information.

Security Reliability and Future AI Roadmap

Secure Local AI Inference Preventing Data Leaks

Data privacy is a paramount concern for Japanese corporations, especially given the strict requirements of the Japanese Act on the Protection of Personal Information, which saw significant amendments on April 7, 2026. Qsirch 7.1.0 addresses these concerns by performing all AI inference processes—including Embedding, ASR (Automatic Speech Recognition), VLM, and LLM—entirely within the NAS device. This “complete local inference” ensures that sensitive corporate data or personal information is never transmitted to external cloud servers or third-party providers. This makes the system ideal for secure environments, such as medical, financial, and legal sectors in Japan, where air-gapped or strictly controlled intranet operations are required.

RAG Integration and Conversational AI Capabilities

QNAP’s roadmap for Qsirch involves evolving the tool into a sophisticated AI assistant. Future updates, such as Qsirch 6.0, aim to integrate Retrieval-Augmented Generation (RAG) with on-premise models like DeepSeek or Gemma. This will allow users to have multi-turn conversations with the AI, asking complex questions based on the private data stored within their NAS. Another innovative feature is the AI Summary, which generates a concise summary of documents, images, videos, or emails with a single click. To support global business operations, these summaries are automatically translated into the user’s NAS interface language, facilitating quick information gathering across different languages. The following image illustrates the AI Summary interface. .

System Optimization and Hardware Requirements

While AI features offer great benefits, they can place a heavy load on system resources. Based on feedback from beta testers who experienced system lag during indexing, QNAP has introduced Smart Indexing Scheduling. This allows the AI to perform resource-intensive indexing during off-peak hours, such as late at night, ensuring that NAS performance remains optimal during working hours. To utilize AI Mode effectively, QNAP recommends high-performance hardware, specifically the x74 series or higher, equipped with a GPU containing at least 16GB of VRAM. However, the standard keyword search remains available on all compatible models without the need for a dedicated GPU, allowing for flexible operation depending on the hardware configuration.

Reference Pages

• [QNAP Qsirch Integrates On-Premise VLM and LLM AI to Enhance Cross-Media Semantic Search on NAS] https://www.qnap.com/en/news/2026/qnap-qsirch-integrates-on-premise-vlm-and-llm-ai-to-enhance-cross-media-semantic-search-on-nas

• [QNAP Makes All Qsirch Premium Features Free, Announces AI-Powered Search Enhancements] https://www.qnap.com/en/news/2025/qnap-makes-all-qsirch-premium-features-free-announces-ai-powered-search-enhancements

• [Qsirch App Release Notes | QNAP] https://www.qnap.com/en/app-release-notes?keyword=Qsi&app=Qsirch&product-line=nas

【#QNAP #NAS #Qsirch #ArtificialIntelligence #SemanticSearch #OnPremiseAI #DataManagement #TechNews】