Japanese Developer Ryoma Nakajima Releases Open-Source AI Video Editor 'AKARI Video' v1.0.0

Japanese engineer Ryoma Nakajima officially released AI-native video editor AKARI Video v1.0.0 on October 2, 2026. The MIT-licensed open-source tool connects to

tau · October 3, 2026

#AKARI Video #AI Video Editing #Open Source #Claude Code #Codex #AI Agents

Japanese Developer Ryoma Nakajima Releases Open-Source AI Video Editor 'AKARI Video' v1.0.0

Japanese developer Ryoma Nakajima (@ryoma_nakajima) officially released version 1.0.0 of "AKARI Video," an AI-native video editing application, on October 2, 2026. Designed as a desktop application where users can initiate video editing workflows simply by speaking in Japanese or natural instructions, the tool is distributed as an open-source project that is free to start using.

AKARI Video AI video editor interface showing timeline and natural language editing assistant

Image source: @ryoma_nakajima

The v1.0.0 release is tailored for creators who find conventional video editing software complex or time-consuming. According to the developer, users do not need to memorize complicated editing shortcuts; instead, they convey their goals in natural language, while the AI agent handles everything from transcription to text overlay placement, leaving the creator to review and adjust only the final details.

Agent Connectivity and Operating Models

At the core of AKARI Video is its flexible AI agent integration. Creators can connect existing accounts for ChatGPT, Claude, or Google to run Claude Code, Codex, or Antigravity (Gemini) as their editing assistant. The platform also supports free-tier AI models for users without active paid subscriptions.

According to the official project site, desktop builds are available for macOS (Apple Silicon) and Windows (x64). For developers and power users who already operate in a terminal, AKARI Video provides a one-line CLI installation path that sets up the full runtime under ~/.akari/app. The codebase is hosted publicly on GitHub under the AkariLabs/akari-video repository with an MIT license.

Users should note that while the application itself is open source and free, connecting high-tier frontier models like Claude Code or Codex requires the user's own subscription or API credentials (typically starting around $20 per month), though free alternative AI partners can also be selected.

7,000+ Asset Library and Automated Transcription

Alongside the application, Nakajima introduced "Asset Lab" (素材ラボ), a library containing over 7,000 free, commercially usable assets including text overlays, background music, and illustrations. The developer notes that the AI understands the context and tone of the footage—for example, detecting a melancholy scene and automatically selecting a fitting subtitle style and subdued audio accompaniment.

For speech-to-text processing, Nakajima confirmed in replies that AKARI Video can utilize its built-in local Whisper engine or external transcription APIs. This enables solo creators to produce subtitle-ready social media content using only their raw footage and recorded remarks.

Availability, Download Paths, and Early Cautions

Installers are available directly from the official download portal (akari.video/download), offering .dmg and .zip archives for macOS alongside .exe installers and portable zip builds for 64-bit Windows. Nakajima shared that the tool has undergone more than 100 iterative updates during development, with feature requests and bug reports managed via Discord.

On Windows, initial launches may trigger a blue Windows SmartScreen warning because code signing is still in progress. The official documentation instructs users to proceed via "More info" → "Run anyway," with the developer noting that official signing certification is expected later this month. While the core software and 7,000 initial assets are free, supplementary paid asset packs and a 500-seat lifetime pass are offered separately as commercial add-ons.

Sources