Ghost Hand Editing

What is ghost hand editing?

GhostCut, whose English name is GhostCut, is a one-stop AI subtitling platform designed for using in short-form videos, e-commerce, social media, educational courses, and marketing videos aimed at international audiences. It integrates subtitle recognition and timing adjustment, translation and proofreading, removal of original subtitles, tagging of multiple characters, AI voice synthesis, voice cloning, background music processing, and video rendering into a single online workflow, thereby reducing the need to repeatedly import and export files between different software tools for transcription, translation, audio processing, and editing.

Guishou Editing can be more accurately described as a \"tool for the localization of AI videos and their batch translation\", rather than a general video generation model that creates complete videos directly from textual prompts. It is capable of handling individual videos created by creators, and it also offers project management functions, as well as the ability to process hundreds or even thousands of video clips at once through APIs; it is thus suitable for teams that need to deliver content in multiple languages on a large scale.

Core functions

1. AI video translation and dubbing

Users can upload local videos or paste video links from supported social platforms; after selecting the source language, target language, subtitle options, voiceover, and background music, they submit the file. The system uses ASR to extract the spoken audio, carries out the translation, adds an AI voiceover, and synchronizes the audio with the visuals. It is also possible to remove existing subtitles and add new ones within the same task.

The current product page states that it supports over 40 target languages, with voiceovers available in dozens of languages and regions including those in Europe, the Americas, Japan, Korea, and Southeast Asia. The availability of different languages, voice tones, and functions varies; before starting mass production, it is necessary to conduct sample tests using actual content to check accents, proper nouns, tone, and timing.

2. Subtitle extraction, alignment, and translation

The platform combines ASR and OCR: ASR is suitable for extracting dialogue from clear spoken voices, while OCR is used to identify text subtitles in images; it is particularly useful for videos that lack clear spoken audio or where the original subtitle information needs to be retained. The official website states that recognition is supported in over 100 languages, and the system has been optimized to handle noise, visual disturbances, and batch processing of multiple files.

Automatic recognition may still lead to errors in identifying names, brands, numbers, dialects, and background noises. Important content should be reviewed in the online editor, checking both the original text, the translation, and the timeline, before moving on to the voice-over and rendering stages, in order to prevent such recognition errors from worsening over time.

3. Translation review for large models

Guishou Editing currently uses models such as DeepSeek, along with a multi-Agent translation and review process, to focus on ensuring consistency in context, terminology, and the naturalness of expression. For short dramas and series, the platform also emphasizes multi-modal recognition of characters, voice signatures, and text across different episodes, in order to avoid inconsistencies in a character’s name or voice from one episode to another.

AI proofreading cannot replace native-speaking translators. Marketing promises, key plot points, legal terminology, medical content, cultural taboos, and terms that are sensitive on certain platforms need to be reviewed by people from the target market.

The humor, irony, and character relationships in the same sentence can also be misinterpreted due to the lack of contextual information about the plot.

4. Traceless subtitles, removal of text and watermarks

It can automatically detect and remove hard subtitles, name tags, titles, text watermarks, and logos; it also uses image restoration models to complete the background. The Pro mode allows users to select specific areas, objects, and time ranges, making it suitable for moving elements or dealing with complex images.

Subtitle removal can be submitted simultaneously with video translation, and batch processing of up to 100 videos at a time is also supported.

The effectiveness of erasure is influenced by factors such as the degree of movement, transparency, complexity of textures, occlusion, and camera changes; in detailed areas, smearing, flickering, or errors in background reconstruction may occur. More importantly, just because it is technically possible to erase something does not mean that it is permissible to remove someone else’s identifying details or reuse their work – the user must possess the copyright to that material or have obtained authorization.

5. Multi-role recognition, AI voice synthesis, and voice cloning

The professional translation process automatically extracts the dialogue and identifies the different speakers. Users can adjust the character labels online, select different voice tones for each character, and then generate the dubbed version. The platform offers various quality levels such as classic, ultra-realistic, and highly expressive cloning; it also allows keeping the background sounds, controlling the original audio track, and creating multiple versions with different characters.

Voice cloning can only be used with one’s own voice or a voice for which explicit permission has been obtained. It is not allowed to impersonate others, forge endorsements, create fraudulent audio messages, or bypass the platform’s identification requirements.

When dealing with the voices of actors, hosts, employees, and customers, it is necessary to specify the scope of authorization, the duration of use, and the mechanisms for withdrawal.

6. Online subtitle editor

The free online subtitle editor allows for simultaneous viewing of the original text and its translation, editing of subtitle content and the timeline, batch translation of SRT files, and automatic labeling of speakers. After proofreading is complete, further work such as adding voiceovers, deleting content, and rendering can be carried out; subtitle files such as SRT can also be downloaded.

The length of the subtitles should be adjusted to match the reading speed of the target language. When Chinese text is translated into languages such as English or Spanish, it usually becomes longer; therefore, manual adjustments are needed regarding sentence breaks, the number of characters per line, and the pause time, so as to prevent the text from covering the screen or preventing viewers from having enough time to read it.

7. Background music and track processing

For video translation, it is possible to choose the original audio with silence, to keep the background music and non-human sound effects, or to have AI generate an appropriate soundtrack. The intelligent soundtrack analysis the changes in the visuals and the rhythm, and selects music from royalty-free tracks;

The platform also offers alternatives for background music that might pose copyright issues.

\"Royalty-free\" does not mean that there are no licensing conditions. Before uploading music to YouTube, TikTok, advertising platforms, or customer channels, it is necessary to ensure that the music license covers commercial use, the relevant regions, the duration of use, and the account in question, and to keep the authorization documents on hand.

8. AI commentary and video reconstruction

AI commentary can understand uploaded movie clips, interviews, variety shows, or short dramas, automatically generate commentary scripts and carry out basic editing. Video reconstruction is intended for long interviews, live broadcast replays, and existing marketing videos; it allows for the extraction of key moments, rewriting of text, and creation of segments suitable for short-video platforms.

These functions are suitable for creating drafts and testing various versions; copyrighted video clips must not be used without permission for commentary, distribution, or commercial purposes. Selection of key moments, presentation of facts, editing semantics, and the scope of citations still require manual review.

9. Image translation and text removal

In addition to videos, GhostHand Editing also offers services for translating images used in e-commerce, as well as for automatically removing or repairing text from images. Image translation can be used to localize product images on platforms such as Amazon, Shopify, eBay, and TikTok.

Erasure supports detection of over 100 languages, as well as text that is tilted or written vertically, and it also attempts to preserve the text on the product itself.

Automatic filling-in may alter font sizes, line breaks, brand names, and product labels. Before listing the product, it is necessary to verify its specifications, price, compliance statements, and packaging details; the generated content must not be used to conceal the actual product or to create fake certifications.

Project management and batch processing

The project workspace is used to manage videos, subtitles, audio, and multilingual final products; it enables the simultaneous uploading, processing, and translation of large amounts of material. The official website’s documentation states that up to 100 videos can be processed at a time for certain tasks, while corporate translation processes allow for the management of even larger volumes of content on a project basis.

Batch tasks consume points rapidly, and a mistake in the subtitles of one language version can spread to all other language versions as well. The team should first create a glossary, a list of characters, and criteria for quality checks; after verifying with a small sample, they can submit the entire batch, while keeping track of the correspondence between the original footage, the subtitles, and the final version.

Exporting engineering files

After the translation is complete, it is possible to preview and download the video without any platform watermark; it is also possible to download individual files such as SRT subtitles, audio, and music. The official website also states that project files from apps like CapCut can be exported, which facilitates further adjustments to subtitle styles, shots, volume, and the overall formatting of the video in professional editing software.

The compatibility of engineering files can be affected by the software version, fonts, audio license rights, and the path to local assets. It is necessary to reopen the files in the target software before delivery for verification, and to ensure that fonts, audio files, and other assets are properly licensed.

API and enterprise integration

Guishou’s editing API allows integration of features such as silent subtitle removal, video translation, one-click narration, image translation, and image deletion. The steps for integration provided on the website’s member page include activating the service, obtaining keys, purchasing credit packs, and integrating the code; the API uses the same credit-based billing logic as the website version.

APIs are suitable for content platforms, short-form drama distribution systems, and e-commerce asset processing pipelines. Integrators are responsible for protecting the keys, handling uploads, callbacks, retry attempts in case of failures, ensuring idempotency, managing usage alerts and cleaning up results, as well as confirming parameters such as concurrency levels, file size limits, data storage locations, and the terms of the enterprise service agreement.

Free version, membership, and point prices

Guishou Editing offers a free trial version, various membership levels, prepaid card options, as well as corporate solutions. The free account allows up to 15 seconds per video, 400 MB per file, 100 MB of storage space in the cloud, and storage of videos for 7 days.

Different payment tiers increase the maximum duration of a single video to 6 or 15 minutes, raise the file size limit to 1GB, extend the storage period to 30 days, and provide better point discounts.

The specific grade names, amounts, and promotional discounts are subject to those listed on the purchase page after logging in.

The platform charges based on the functions used, with a billing unit of 30 seconds. If a video is shorter than 30 seconds, it is still counted as 30 seconds. When basic processing, translation, dubbing, and caption removal are all used simultaneously, the costs for each service are added together. According to the current list of available benefits, the cost for caption removal in the Lite version starts at 4 points per 30 seconds; different membership levels can reduce this cost to 2 points, 1.5 points, etc.

The erasure function via the Pro option starts at around 6 points per 30 seconds; classic AI voiceovers start at around 2 points, ultra-realistic voiceovers start at around 50 points, and high-emotion cloning starts at around 45 points.

For corporate accounts, please contact the business department.

The official website’s marketing page indicates that the cost for translation and voice-over services can be as low as 0.2 yuan per minute, while the initial version of the page showed a minimum cost of 0.5 yuan per minute. Such low prices depend on the package chosen, any point discounts, and the combination of features offered; they do not represent a uniform price for all translation, editing, and voice-over tasks.

Before making a purchase, it is necessary to calculate the cost in the billing details based on the actual duration and the combination of features used.

The point rules also state that points that have expired within 30 days can be automatically recovered after a purchase, but such recovered points are not eligible for a refund. Membership does not mean unlimited access to all features; high-definition videos, duration limits, concurrent usage, batch processing, API access, and advanced sound effects may still be subject to point or level restrictions.

Supported platforms

• On the web version, it can be used in browsers on Windows and macOS.
• Mobile browser H5;
• WeChat Mini Programs;
• The iPhone and Android can be used via mobile web pages or mini-programs, while complex batch tasks are better handled on a computer.
• Integration of enterprise APIs with content systems.

Company, Privacy, and Open Source Status

The policy regarding the protection of personal information in connection with the Ghost Hand editing tool states that the company responsible for its operation is Shanghai Zhaoli Technology Co., Ltd. This platform offers commercial cloud-based services, and it does not make the complete source code for its video recognition, translation, deletion, dubbing, and rendering functions available, which means it is not an open-source video tool.

Providing an API does not equate to making the source code available or allowing private copying.

Videos, audio, subtitles, and images need to be uploaded to the cloud for processing. The official website states that data is stored and backed up in multiple cloud environments around the world, with encryption, data isolation, and access control in place.

Enterprise clients should still, in accordance with the contract, determine the retention period for data, the methods of deletion, cross-border data transfer, usage for model training, employee access rights, and procedures for notifying in case of incidents.

Confidential materials, unreleased episodes, and personal biometric voices should not be uploaded without the organization’s approval.

Guide to using Ghost Hand editing

Complete a basic task.

1. Identify the audience, platform, format, duration, and the information that needs to be conveyed;
2. In Ghost Hand editing, prepare scripts, footage, or reference materials that have permission to be used;
3. Select AI video translation and dubbing to create a low-cost preview;
4. Use subtitle extraction, framing, and translation to adjust the visuals, rhythm, subtitles, and audio;
5. Check each frame for characters, text, logos, lip movements, and factual accuracy;
6. Export in the desired format after confirming the licensing for music, portraits, and materials;

Create reusable professional workflows

1. Create scripts, shot lists, brand assets, and a list of elements that are prohibited from use;
2. Unified parameters are set for AI video translation and dubbing, subtitle extraction, framing and translation, as well as review and saving of translations using large models.
3. First, use representative shots to test the model and the quota;
4. Transfer the failed shots to manual editing or regenerate them;
5. Uniformize subtitles, volume, colors, and end credits;
6. Record the version and reviewer before publishing in batches;

Which users is it suitable for?

• A distribution team that translates short dramas, web comics, and humanoid dramas into multiple language versions on a bulk basis;
• E-commerce companies that localize marketing materials for platforms such as TikTok and YouTube;
• Bloggers and MCNs that create subtitles and voiceovers for overseas audiences;
• Educational institutions that offer translation courses, training, and knowledge videos;
• A film and television localization team that handles multiple character voices, audio, and subtitles;
• Companies that build automated video processing pipelines using APIs.

Product advantages

• It covers the entire translation and production process, including extraction, translation, proofreading, deletion, voice-over, music addition, and synthesis;
• Use both ASR and OCR to handle both spoken audio and textual subtitles on the screen;
• It supports multi-role recognition, voice cloning, and consistency handling across different sets;
• It supports batch processing of up to 100 videos and provides project management;
• Supports online subtitle proofreading, as well as export in SRT and project format;
• Both video and image functions can be integrated through APIs.

Restrictions and Precautions

• Video translation and dubbing are affected by the quality of the original audio, speaking speed, dialects, background music, and overlapping voices; the automatic results are not guaranteed to be entirely accurate.
• Before official release, it is necessary to check the subtitle timeline, translations, characters, voice tones, lip synchronization, and the alignment between audio and visuals.
• Subtitle removal, watermark removal, commentary, and remastering shall not be used for unauthorized distribution, fake original creations, or circumventing copyright protections;
• Voice cloning must not be used to impersonate real people;
• Users must comply simultaneously with the laws of the platform from which the material originates, the platform on which it is to be published, and the jurisdiction in which they are located;
• The charge is calculated on a 30-second basis, with additional fees added for multiple tasks being processed.
• The “lowest” price is not a fixed order price;
• Before handling large-scale tasks, it is necessary to estimate the number of points required, and to determine which tasks will fail, what is the expiration date for those points, what is the storage period, and what are the rules regarding refunds.

Frequently Asked Questions

Can Ghost Hand editing be used for free?

It is possible to experience some basic functions, but free accounts have limited limits on video length, file size, storage space, retention time, and the number of points available. Advanced features such as translation, deletion, dubbing, voice cloning, batch processing, and API access typically require a membership or points.

How many languages can Ghost Hand Editing translate videos into?

The current video translation page claims to support over 40 target languages, with a wider range of subtitle recognition options. The level of support for recognition, translation, and dubbing may vary depending on the language; it is necessary to check this on the task page.

Will there be a watermark on the exported video?

The official website states that both free and paid versions of the exported subtitled videos do not contain any watermarks from unauthorized editing tools; however, to remove the identifiers from the original video, a legitimate license is required.

Can it handle multi-character short plays?

Yes. The professional version of the system can automatically identify characters, proofread dialogue, select voices for different characters, and generate multi-character voiceovers; it is recommended to conduct a manual review to ensure consistency of characters across different episodes.

How much does Ghost Hand Editing charge?

Free trials, membership plans, prepaid card options, and enterprise solutions are available. Charges are calculated based on tasks performed in 30-second intervals, with multiple functions resulting in a cumulative fee.

The final amount should be calculated based on the price per click on the purchase page and the task combination.

Does GhostHand Editing provide an API?

It offers functions such as removing subtitles from videos, providing translations and commentary, as well as translating and removing images. The pricing logic is the same for both the API and the web interface; key details regarding keys, concurrency limits, and enterprise-related terms need to be confirmed separately.

Is GhostHand Editing open-source software?

No. It is a commercial cloud-based video translation platform; it offers APIs, but the complete source code of its core system is not made public.