Hachi

Natural language search for your personal videos and images.

Intermediate
Screenshot of Hachi, Natural language search for your personal videos and images

What is Hachi?

Hachi is a self-hosted, offline AI tool for natural language search across personal videos and images. It indexes your media locally, supports face recognition, and runs on consumer-grade CPUs without sending data to the cloud. A free version is available on GitHub, with a paid upgrade for face search and support.

Hachi is a self-hosted web app that lets you search your personal videos and images using natural language. Instead of manually scrolling through folders or tagging files by hand, you type a description of what you are looking for and Hachi returns the matching scenes, objects, or people. It also includes face recognition that can find a specific person across an entire image collection. The tool runs entirely on your own machine, with no internet connection required, so your media never leaves your device. Under the hood, Hachi builds a local index of your videos and images the first time you add them. Once indexed, searches are fast and work across the whole library, not just one file at a time. For videos, you can search while playback is running, which is useful for finding a particular moment in a long recording. The face recognition feature is also fully local, letting you search for yourself or loved ones without uploading anything to a server. The product is aimed at individuals and small teams who want a private, offline way to manage large media collections. It suits photographers, videographers, researchers, and anyone who values data privacy. Because it is self-hosted, you need a machine with a quad-core CPU supporting AVX2 instructions and at least 4GB of RAM, running an x86_64 architecture. Setup is more involved than a typical cloud service, but the trade-off is complete control over your data. Hachi is open source under the AGPLv3, with a free version available on GitHub and a paid version sold through Gumroad. The free tier covers live video search, natural language search across indexed videos and images, and unlimited indexing. The paid version adds face recognition across indexed images, improved video search, email support, and one year of free updates. The paid license is for personal, non-commercial use only. Hachi competes with cloud-based media search tools by offering privacy and offline operation, though it requires more technical setup and lacks some polish of commercial alternatives.

How do you use Hachi?

  1. 1Check system requirements
    Confirm your computer has a quad-core x86_64 processor with AVX2 support and at least 4GB of RAM. The hardware requirements page lists the details.
    Hachi — Check system requirements
  2. 2Download and install Hachi
    Get the free version from the GitHub repository or purchase the paid version on Gumroad. Follow the installation instructions for your platform.
  3. 3Launch the local web app
    Start Hachi on your machine and open the web interface in your browser. No internet connection is needed after installation.
  4. 4Index your media library
    Add folders containing your videos and images. Hachi builds a local index the first time, which may take some time depending on library size.
  5. 5Search with natural language
    Type descriptive queries to find scenes, objects, or people across all indexed media. Use face recognition in the paid version to search by a specific face.

Pros and cons

Pros

  • Works entirely offline, so your media never leaves your device.AI
  • Natural language search across both videos and images in one interface.AI
  • Face recognition is fully local and helps find specific people quickly.AI
  • Indexes media only once, making repeated searches fast.AI
  • Open source core (AGPLv3) with a free version to try.AI

Cons

  • Requires a quad-core CPU with AVX2 and at least 4GB RAM, which excludes some older machines.AI
  • Only supports x86_64 architectures, so ARM devices like Apple Silicon are not covered.AI
  • Setup is self-hosted, which may be too technical for non-developers.AI
  • Paid version is limited to personal, non-commercial use only.AI
  • No mobile or cloud version, and no API is mentioned.AI

How much does Hachi cost?

Free trial

A free version is available on GitHub, offering core search features without a time limit. The paid version does not mention a free trial.

Learn more

Pricing

Hachi has a free version available on GitHub with live video search, natural language search across indexed videos and images, and unlimited indexing. The paid version, sold on Gumroad, adds face recognition across images, improved video search, email support, and one year of free updates. Pricing details are not listed on the homepage.

Learn more

Support

The free version includes community support, while the paid version adds email support. No further support details are provided.

Features

Hachi offers natural language search across videos and images, including live search while a video plays. It indexes media locally for fast repeated searches, supports face recognition for finding people across image collections, and provides full database search across all indexed files. All processing is offline and private.

Frequently asked questions about Hachi

  • How much does Hachi cost?
    Hachi has a free version available on GitHub with live video search, natural language search across indexed videos and images, and unlimited indexing. The paid version, sold on Gumroad, adds face recognition across images, improved video search, email support, and one year of free updates. Pricing details are not listed on the homepage.
  • Does Hachi offer a free trial?
    A free version is available on GitHub, offering core search features without a time limit. The paid version does not mention a free trial.
  • How do you use Hachi?
    The walkthrough on this page covers 5 steps: 1. Check system requirements 2. Download and install Hachi 3. Launch the local web app 4. Index your media library 5. Search with natural language.
  • What platforms does Hachi support?
    Hachi is available on Web App, Linux, Windows and MacOS.
  • What are the limitations of Hachi?
    Requires a quad-core CPU with AVX2 and at least 4GB RAM, which excludes some older machines. Only supports x86_64 architectures, so ARM devices like Apple Silicon are not covered. Setup is self-hosted, which may be too technical for non-developers.

Status

StatusActive
Views0
Outbound clicks0
Added10/21/2026

Platforms

Web AppLinuxWindowsMacOS

Pricing

FreePaid