Mavericks Inc. Revamps Editor for AI Video Generation Tool 'NoLang', Integrating Creation and Finishing in a Single Interface for Intuitive Drag-and-Drop Editing That Reduces Video Production Time

Mavericks Inc. (Headquarters: Bunkyo City, Tokyo; Representative Director and President: Shotaro Okuno), provider of the AI video generation tool 'NoLang', has completely overhauled the NoLang editor interface. Elements such as background, images, video, text, BGM, sound effects, characters, and subtitles are now separated into individual lanes on a timeline, allowing users to directly adjust their position and duration via drag-and-drop operations. The entire workflow—from AI generation to final editing on the timeline—can now be completed within a single screen, eliminating the need for repeated AI regeneration and enabling users to make precise, targeted edits to specific sections.

The Weakness of Generative AI: Minor Edits Can Alter the Entire Output

In environments where generative AI is used to create images and videos, significant time is often spent on revisions before completion. According to a survey by PRIZMA Inc. (conducted March 6–8, 2026) of 1,003 individuals with professional experience using generative AI for image and video production, the most common complaint (44.2%) was that 'when trying to fix just one part, the instructions cause other well-made sections to disappear.' The number of revisions required reached 60.8% for 'four or more times' (30.5% for 4–6 times, 16.5% for 7–9 times, and 13.8% for 10+ times), and 76.4% of respondents reported feeling 'burdened' by the revision process (20.1% 'strongly,' 56.3% 'somewhat'). The most common response for time spent per revision was 'less than 1 to 3 hours' at 27.8% (※1). Significant time is consumed in the back-and-forth process of regenerating the video after giving natural language instructions to correct sections that don't meet expectations.

When users want to make only minor adjustments, AI generation processes—designed as batch operations—make it difficult to precisely target only the intended differences. The amount of information that can be conveyed through prompts is limited, and uncommunicated parts are often regenerated unnecessarily. In response to the demand for a tool that enables 'immediate editing of only the intended sections' after 'initial draft generation by AI,' Mavericks has completely overhauled the editor interface of its AI video generation tool 'NoLang.'

※1 PRIZMA Inc., 'Survey on the Realities of Revisions and Fine-Tuning in Image and Video Production Using Generative AI' (March 6–8, 2026; survey of 1,003 men and women aged 20–50 with daily professional experience using generative AI for image and video production)

Revamped NoLang Editor: Simplifying Post-Generation Editing

In the updated editor, elements such as background, images, video, text, BGM, sound effects, characters, and subtitles are now separated into independent lanes, allowing them to be viewed and edited along a timeline. The interface, previously used only to preview generated results, has now become an editable timeline where direct adjustments can be made.

While NoLang previously offered an editing interface alongside its automatic video generation from materials such as text, PDFs, and URLs, the range of adjustable parameters was limited. With this update, each element—background, images, video, text, BGM, sound effects, characters, and subtitles—can now be placed on separate lanes and directly adjusted on the timeline, greatly expanding editing flexibility. Users can open and edit previously generated video projects in the same timeline interface, eliminating the need to recreate videos made before the update.

Intuitive Usability, Even for Those Without Video Editing Experience

On the timeline lanes, users can drag elements such as background, images, video, text, BGM, and sound effects to adjust their position and duration on the timeline. Right-clicking a segment (individual components of the video) displays an editing menu with options such as 'Split,' 'Extend Forward,' and 'Extend Backward,' enabling trimming and splitting through on-screen operations alone—without needing to memorize shortcuts or navigate complex menu hierarchies of professional software. Selecting 'Extend Forward' instantly expands the boundary to just before the adjacent segment, or to the beginning of the lane if no adjacent segment exists, eliminating the need for repeated, frame-by-frame drag operations.

Each lane is labeled with icons and color-coded, making it immediately clear which element—background, image, video, text, BGM, sound effects, character, or subtitle—can be adjusted where. According to a survey by Nile Co., Ltd. on 'Appliv' (conducted November 13–20, 2023, targeting 557 video editing app users aged 10–60), the top user concern was 'complex operation methods,' cited by approximately 41% (229 people) (※2). NoLang addresses this issue with an interface design that does not assume familiarity with video editing software terminology or complex menu systems. For example, users looking to edit subtitle text go to the text lane, and those adjusting BGM timing go to the BGM lane—the color-coding directly links the task to the on-screen location, eliminating the need to navigate through hierarchical menus to find the desired function.

Additionally, the timeline and preview screen are synchronized, so adjustments made on the lanes are immediately reflected in the preview. There is no need to re-export the video to check changes, such as rewritten subtitle text. The ability to see the results of operations in real time further enhances usability for users without prior video editing experience.

※2 Nile Co., Ltd., 'Survey on Video Editing Apps' (Appliv survey, conducted November 13–20, 2023, targeting 557 video editing app users aged 10–60)

Completing the Entire Video Workflow—from Generation to Final Output—Within a Single NoLang Interface

Direct editing on these lanes is available immediately after AI generation. Once the AI generates the video foundation, users can transition directly to the timeline without switching screens and adjust any sections of concern.

Traditional video production has historically involved two types of 'back-and-forth' processes. The first is the cycle of repeatedly instructing the AI to regenerate sections that don't meet expectations using natural language. As noted in the PRIZMA survey, 60.8% of users required four or more revisions; NoLang eliminates this cycle by enabling direct editing of specific sections on the timeline instead of repeated regeneration.

The second is the process of exporting the AI-generated video to another video editing software, making finishing adjustments there, and then re-exporting. With NoLang, the generated video is directly editable on the timeline, eliminating the need to export files and reopen them in separate software.

The workflow of 'letting AI handle the initial draft and finishing it manually' can now be completed entirely within NoLang, without switching between software. An editing environment that allows users to precisely target only the intended sections—rather than repeatedly rewriting prompts to find the right output—reduces the number of back-and-forth cycles.

Future Outlook

Mavericks plans to refine usability further, starting from this overhaul that integrates generation and timeline editing into a seamless workflow. Development will continue not only on expanding generative AI capabilities but also on editing functions that handle the 'finishing' stage, with a focus on intuitive, hassle-free operability.

About the AI Video Generation Tool 'NoLang'

The AI video generation tool 'NoLang' automatically creates videos from text, PDFs, PowerPoint (pptx) files, and website URLs.