Create a frame per second

What is a frame per second?

FrameSec, also known as SecCreate, is an AI-based content creation and online video production platform developed by Xinyi (Beijing) Technology Co., Ltd. Built on Xinyi’s video large model and the SecCreate AIGC intelligent engine, this platform integrates text generation, semantic storyboarding, material retrieval, AI voiceovers, subtitles, digital avatars, image creation, and video synthesis into a single web-based workflow.

It is primarily aimed at individuals in the field of social media, corporate operations, integrated media, knowledge bloggers, and marketing teams who lack professional editing experience. The core function of this platform is not to generate each frame from scratch, but rather to analyze the text content, automatically divide it into segments, and then match appropriate visuals from a library or from the users’ own materials. Together with sound, subtitles, templates, and other elements, this allows for the quick creation of finished videos.

AI conversion of text and images to video

Converting text and images into videos is the most representative feature of Secreat. Users can enter text, paste article content, import compatible article links, or upload Word documents; the system automatically analyzes the title and main text, divides the content into separate segments based on semantic meaning, and matches video or image materials for each segment of text.

It automatically matches content with a clear structure, such as news, knowledge, popular science articles, emotional content, historical material, lifestyle topics, and product descriptions. For abstract concepts, specialized equipment, lesser-known individuals, and stories with strong narratives, the available materials may be inaccurate or repetitive; in such cases, manual search, replacement, and adjustment of the duration are required.

Article link and Word import

The official help documentation states that the tool for converting text and images into videos allows for direct import of text, import via article links, and creation of videos from Word documents. Using links eliminates the need to copy and paste text, while the Word method is suitable for longer texts; it also helps to retain some of the original images. Once the document is opened for editing, it can still be reorganized or modified.

The page structures of platforms such as WeChat Official Accounts, REDnote, Baidu Baijia, Toutiao, Sohu Accounts, and Weibo keep changing, and not all links can be resolved reliably. When faced with issues related to logging in, anti-scraping measures, permissions, or dynamic loading, it is better to paste the text or import it from Word, while also ensuring that permission for reproduction has been obtained.

AI-powered scene segmentation and asset matching

MiChuang performs semantic analysis on the text, breaking down long texts into sentences and phrases that are suitable for use in videos; it then searches for relevant materials based on keywords and industry specifics. Users can modify the segments, keywords, visuals, order, and duration, thereby avoiding complete reliance on the results generated initially.

The more specific the keywords are, the more accurate the matching will usually be. Factual elements such as names of people, locations, dates, and product models need to be verified manually; it is not acceptable to use seemingly similar materials in place of actual events, as this can lead to misrepresentation.

Horizontal and vertical videos

When creating a project, you can choose between horizontal or vertical templates that suit common aspect ratios such as 16:9 and 9:16. The platform also offers an intelligent tool for converting horizontal videos to vertical formats with ratios of 9:16 or 3:4, thereby reducing the need to re-edit them for different platforms.

Automatic cropping may result in the removal of characters, subtitles, products, or the presentation interface. After conversion, it is necessary to check each frame for proper subject tracking, subtitle placement, the logo, the cover image, and the dimensions recommended by the platform.

AI helps with writing

AI-assisted writing allows for the creation of text based on titles or questions; it can also rewrite, continue, refine, and expand content, covering areas such as social media posts, industry-related content, SEO optimization, and video scripts. The generated text can be selected directly to proceed with the AI video creation process, thus reducing the need to copy content between different tools.

AI-generated copy may contain factual errors, repetitions, vagueness, and issues related to copyright. Information on topics such as news, history, health, law, finance, and product specifications must be verified; when content is published in large quantities, it is important to avoid high levels of homogeneity in order to maintain the quality of the platform.

AI voice synthesis and text-to-speech

Video editors offer a variety of AI voices in Chinese and foreign languages; these voices can be filtered by gender, language, scenario, and style. It is also possible to listen to them, adjust the speaking speed, and replace the voice tone. There is also a separate text-to-speech tool that allows scripts to be converted into audio content for use in short videos, courses, and audio materials.

The platform supports multiple pronunciations for characters, digital readings, and adjustments to sentence punctuation. For names, places, brands, years, units, and English abbreviations, it is still necessary to listen to each phrase individually;

It is best to create long texts in paragraphs, so as to avoid having to rewrite an entire paragraph each time a change is made.

AI digital human broadcasting

The digital avatar module offers over 100 realistic character models representing various professions and coming from different countries and regions; it enables the combination of text content with AI-generated voice to create video narrations. It is suitable for providing knowledge explanations, product introductions, corporate training, store promotions, as well as for situations where it is not possible to use a real person in the videos.

Users usually need to select an avatar, video aspect ratio, voice narration, and background, after which they can generate and download the final video. The avatar design, video duration, clarity, custom avatar options, and live streaming services may fall under different subscription plans or corporate solutions; it is necessary to check the available benefits in real time before making a purchase.

Digital human live streaming and customization

The official digital human product page displays both short videos of digital humans and live streaming services. Companies can inquire about image customization, the creation of exclusive content, and live streaming options; however, details such as delivery times, concurrency levels, latency, driving methods, and prices need to be confirmed through business discussions.

The user agreement specifies that, unless otherwise agreed in writing, customized digital avatar images shall not be used for commercial purposes such as sales, live streaming donations, third-party commercial platforms, advertising, or corporate promotion. The creation of a digital avatar does not imply an automatic grant of commercial rights.

Convert PPT to video

Converting PPTs to videos involves importing the presentation into a platform, where the content of the slides, voiceovers, and digital avatars are used to create explanatory videos. This method is suitable for use in courses, training sessions, product demonstrations, and corporate presentations. Compared to re-editing the material, it allows the existing slide structure to be reused.

Complex animations, embedded fonts, formulas, videos, macros, and special formatting may not be reproduced accurately. After importing, it is necessary to check the page order, text wrapping, proportions, transitions, and the length of the audio narration, to ensure that the information on each page is suitable for viewing in a video format.

AI art generation

AI painting offers various generation models for general images and anime, supporting both text-to-image and image-to-image creation; it can also generate illustrations for texts that lack visual content. The official prompt method involves combining a subject, details, and descriptive words, while users can adjust the artistic style, palette, and reference images.

The images generated may contain errors regarding human anatomy, text, logos, and realistic scenes. When dealing with real people, trademarks, artists’ styles, or commercial uses, it is necessary to check aspects such as portraits, copyright issues, platform policies, and licensing requirements.

Private material library

Users can upload their own videos and images to create a private library of materials. The platform applies intelligent tagging, noise removal, segmentation of video clips, and search functions to the videos, enabling companies or specialized content creators to access their own footage first and addressing the issue of imprecise public materials.

The owner of private materials and the person responsible for granting permissions remains the one who uploaded them. For customer materials, internal products, employee portraits, and unpublished advertisements, clear permissions must be established, and it is necessary to determine whether they will be shared, for how long they will be retained, and how they will be deleted.

Team assets and sharing

Some materials can be shared after going through an approval process, allowing team members with different accounts to reuse the content that has been uploaded. This helps reduce the costs associated with repeated uploading and searching, especially for chain stores, MCNs, and corporate operations.

Before sharing, it is necessary to distinguish between public, team, and private materials in order to prevent unauthorized content from being distributed to other users. When companies require more stringent controls regarding roles, approvals, auditing, and the lifecycle of materials, they should consult enterprise-level solutions.

Templates, subtitles, and brand packaging

The platform allows users to add templates, titles, subtitles, logos, voiceovers, and background music to videos; it also supports vertical formatting. Subtitles can be derived from text or audio, and users can adjust the text, its position, font style, and timing.

Automatic subtitles do not guarantee that all visual elements will be readable. It is necessary to check the number of characters per screen, color contrast, background interference, punctuation, and the safe area for use on mobile devices.

Commercial licenses are also required for the brand logo, fonts, and music.

Video export and project storage

The project is stored in a cloud account, allowing access to it on different computers. Once it has been completed, the video can be submitted for processing and exported; the quality, watermark, duration, and queue priority are determined by the account’s privileges and the current rules of the service.

Cloud-based engineering should not be the only form of backup. Important documents, original materials, voiceovers, subtitles, and the final videos need to be saved locally as well.

It should be exported in advance before the membership expires, the projects are removed, or the platform is updated.

Video widget

Miaochuang offers various tools such as intelligent horizontal-to-vertical conversion, video cropping, watermark removal, and text-to-speech conversion, which can serve as supplements to the main video production process. Cropping supports a range of common aspect ratios, and for watermark removal, it is possible to specify both the time frame and the area to be processed.

Removing watermarks or subtitles is merely a form of image processing; it does not relieve the original video of its copyright restrictions. Permission is still required before using third-party materials in creative works, advertisements, or for commercial purposes.

Material copyright and publication

The platform’s asset library contains content from internal sources, partner sources, or third parties, and the licensing terms for different types of assets may vary. The official help center provides information on whether videos can be published across all platforms as well as details regarding the commercial use of AI-generated voiceovers; it emphasizes that the right to publish a piece of content cannot be determined merely by the fact that it has been successfully exported.

Users should check the identifiers of the materials used at the time of creation, as well as the membership agreement and specific licensing terms, and save the authorization credentials. When searching for materials online, such as video clips, music, fonts, portraits of people, and brand logos, it is particularly important to verify these elements separately.

Members, Points, and Limits

Yi frame per second offers a free trial version, premium membership plans, corporate membership options, as well as other possible subscription packages. Under the paid subscription agreements, members receive points and a certain amount of usage rights; the validity period of these benefits usually coincides with that of the membership itself. The actual privileges are specified on the purchase page.

The current price page uses a dynamic display format, and it is not possible to obtain accurate, up-to-date prices through public data extraction. The monthly fees, discounts, and number of minutes required for production listed in past promotions may no longer be valid, and therefore should not be used as the basis for current pricing.

When making a purchase, it is necessary to check the video length, the number of words in the AI-generated voiceover, the digital avatar, the export options, the clarity, the available materials, and the API quota.

Refunds and purchase channels

Under the payment agreement, members, points, and credit limits are regarded as virtual consumer goods; unless there are legal requirements, provisions specified in the agreement, or significant defects in the service, such items cannot be transferred or refunded after payment has been made. The platform has also issued statements regarding purchase channels, advising users to make purchases through officially approved channels.

Before placing an order, use the free quota to test the suitability of the materials, the virtual characters, the voiceovers, and the export quality; also verify the account details, package options, duration, and auto-renewal rules. Unofficial discounts may not provide any after-sales support.

Enterprise services and APIs

The help center provides information on API interfaces and API billing, targeting organizations that need to generate content in bulk, integrate it into existing systems, or develop custom workflows. Companies can also seek advice regarding private material libraries, digital avatars, media structuring, and solutions for creating industry-specific videos.

The model, concurrency, task duration, callbacks, storage requirements, and pricing related to APIs require login or business approval; the quota available to regular members cannot be considered as the quota for API calls. For production use, authentication, rate limiting, retry mechanisms, as well as the ability to save task status and results are necessary.

Is it open source?

The Yifram Secreat Website, the Xinyi Video Large Model, the Secreat AIGC Engine, and the cloud-based services are not open-source projects; the developers do not provide complete server code that can be deployed locally. Their use requires a platform account, an internet connection, as well as appropriate membership status or quotas.

In the catalog, it should be labeled as a closed-source SaaS solution; the product should not be considered open source just because the underlying scheduling mechanism uses an open-source model. Companies with specific requirements regarding privatization and data residency need to verify these details directly with the vendor.

Tutorial for Creating Frames per Second

Complete a basic task.

1. Identify the audience, platform, format, duration, and the information that needs to be conveyed;
2. Prepare scripts, shots, or reference materials that can be used within a single frame-second creation;
3. Select AI image-to-video conversion to create a low-cost preview;
4. Use the article link along with Word to import and adjust the visuals, pacing, 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. Uniform parameters are established for AI-based conversion of images and text into video, for importing article links and Word documents, as well as for AI-driven scene segmentation and matching of assets.
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?

• Creators who convert official accounts, news articles, educational content, and knowledge-based posts into short videos;
• Social media users who are not familiar with timeline editing but need consistent updates;
• A team responsible for creating digital podcasts, course content, and product explanations;
• Organizations that need to convert PPTs into videos as well as create content for corporate training;
• Companies and MCNs that wish to create private media asset libraries and produce content in bulk.

Product advantages

• Various ways to access content, including text, links, Word documents, and PPT files;
• Semantic storyboarding, material matching, voiceover and subtitle handling, as well as the synthesis process, are all integrated together;
• Digital avatars, AI-assisted writing, and AI-driven drawing can be integrated directly into the final video.
• Supports landscape, portrait, and landscape-to-portrait mode;
• It allows the upload of private materials, along with intelligent tagging and video processing;
• There are abundant Chinese help documents and creation examples.

Restrictions and Precautions

• Automatic video generation is more suitable for structured spoken content, but it does not equate to professional editing;
• The materials, storyboard sequences, voiceovers, subtitles that result from the first match need to be manually modified; they do not correspond to the actual footage.
• It might be difficult to find appropriate content in niche industries.
• Members, points, digital avatars, and APIs come with different quota limits, and past promotions do not reflect the current prices.
• The commercial rights regarding the materials, voiceovers, and digital avatars must be determined in accordance with the terms of the order; in particular, the successful creation of such elements does not imply that they can be used for unlimited commercial purposes across the entire internet.

Frequently Asked Questions

Is One Frame per Second free?

It is possible to register and try out some of the functions, but the length of videos, voiceovers, digital avatars, multimedia materials, export options, and resolution are all limited by the free tier; for frequent use or commercial purposes, a membership or enterprise plan is required.

How much does it cost to create one frame per second?

The official website offers premium membership, corporate membership, as well as a points and quota system; the prices are updated dynamically on the relevant page. The previous monthly fee of 98 yuan or any promotional discounts are no longer valid; the amounts shown on the real-time pricing page should be taken as the reference.

Can articles from a public account be converted into videos?

Videos can be created by using article links, pasting text, or importing from Word; however, link resolution may be affected by page permissions, and permission must be obtained for any content that is reused.

Does Yifram Second Creation have digital humans?

Yes, more than 100 digital avatar designs are available, along with services for short video broadcasts, customization, and live streaming. The scope of commercial use must be determined through a membership agreement or a written contract.

Can the generated video be used for commercial purposes?

It is necessary to check the respective terms regarding members, materials, voiceovers, and digital avatars. The user remains responsible for the copyright of the content they upload as well as any third-party materials; success in exporting the content does not suffice as a criterion.

Is One Frame Per Second open source?

It is not open source. It is an online AI video SaaS service operated by New One Technology, and its core models, engines, and server code are not made public.