JoyAI web version

What is the web version of JoyAI?

The web version of JoyAI is an AIGC creative workshop operated by the JD.com ecosystem, designed for the creation of images, videos, and digital human content. Users can enter prompts in their browser or upload materials to generate and further edit visual content.

It is related to the JoyAI iPhone comprehensive assistant, JoyAI’s enterprise products, as well as the JoyAI model projects made available by JD.com, but it does not represent the same set of functions. This article focuses on the current web-based creation platform; the capabilities of mobile versions or open-source models are not considered part of the web-based capabilities.

Product format and current status

Main functions

AI image generation

Image generation accepts natural language descriptions of scenes, making it suitable for creating concept art, illustrations, visual inspiration, and image captions. The prompts can specify the subject, scene, style, composition, and lighting, and the resulting images are stored as creative assets for future reference.

1. Select the image generation task.
2. It describes the subject and its actions or states.
3. Add background, composition, materials, and colors.
4. Select the model or parameters currently available on the page.
5. Confirm the consumption of points this time.
6. Submit the task and wait for it to be generated.
7. Check the characters, text, hands, and details.
8. Download, copy, or proceed to the editing process.

The generated images may contain spelling errors, structural issues, or stylistic deviations; running the same prompt multiple times can also yield different results. Brand visuals and product presentations need to be manually reviewed, and the example outputs should not be regarded as a guarantee of consistent results.

AI video generation

Video generation allows for the creation of dynamic sequences based on textual descriptions; the platform also offers task types that rely on images, the first and last frames, or multiple images. Users need to describe the subjects in the scene, the movements, any changes in the environment, and the desired outcome, rather than simply providing a general topic.

• Text-to-video: Enter the scene, actions, and camera directions.
• Video from images: Upload a static image and describe the movement you want to see.
• Video with start and end frames: Provides the initial and final images along with an explanation of the transition.
• Multimedia video: Organizes content and motion relationships using multiple reference materials.
• Result: The video file can be previewed and downloaded on the asset page.

The optional modes, video duration, aspect ratio, resolution, and single integration value change according to the settings of the current page. Complex movements, continuous text, consistency in character appearance, and accurate physical processes still require inspection frame by frame.

Digital human video generation

The digital avatar interface accepts the materials related to the character as well as the text that the character is supposed to say, in order to create voiceovers or content featuring virtual characters. Users must first ensure that there is legitimate permission for the character’s appearance, voice, dialogue, and the scenarios in which it is used.

1. Select digital human generation.
2. Select a role provided on the page or upload materials as required.
3. Enter the lines that you want the character to read aloud.
4. Check the name, numbers, brand terms, and pronunciation.
5. Confirm authorization for the portrait, voice, and script.
6. Check the task points and submit them.
7. Check the lip movements, pronunciation, and continuity of the visuals.
8. After downloading, perform the necessary manual editing.

The outputs of digital humans should not be used to impersonate real people, forge identities, or create misinformation. In contexts such as advertising, education, finance, healthcare, and public communications, AI-generated content must also be labeled in accordance with the applicable rules.

Image editing tools

The web page creation process offers various tools for generating and editing images, allowing for further processing of the existing results. The tools currently available include redrawing, removal, intelligent object separation, image enlargement, and clarity enhancement.

The high-definition processing page currently shows examples of fixed points, but the prices for other tools are determined dynamically based on account information. Before submitting a request, it is necessary to refer to the amount indicated next to the button; estimates based on previous screenshots should not be used.

Beta canvas and creative assets

The canvas is used to organize the creation of content within a workspace and to further combine it; it is suitable for exploring various versions or for carrying out sequential creations. The Beta status means that the entry points, nodes, saving methods, and supported formats may continue to change.

The asset creation page serves as a centralized location for storing images and videos created by users, making it easy to view details, continue editing, copy images, or download files. Important finished works should still be saved in storage locations managed by the users themselves.

Agent and Work Exploration

The platform provides information about agents as well as links to the creators’ content, helping users find ideas for creation based on predefined capabilities or existing works. The specific model, permissions, costs, and availability of an agent are determined by the details shown on its page.

Input, output, and upload restrictions

Complete usage tutorial

1. Open the JoyAI web version in your browser.
2. First, browse the creation examples and select a task involving images, videos, or digital avatars.
3. Log in to your JD.com account to claim and use your points.
4. Check the balance of points for the day and the time when they will be reset to zero.
5. Prepare input materials that contain no sensitive information and for which authorization exists.
6. Upload images or enter prompts as required by the task.
7. Select the scale, model, or other parameters provided on the current page.
8. The points consumed are displayed next to the check button.
9. Submit the task and wait for it to be completed.
10. Check for facts, characters, text, actions, and copyright risks.
11. Use redrawing, erasing, image masking, image enlargement, or high-definition processing as needed.
12. Download the finished product and save it in your own project directory.

Prompt writing format

The prompt should separate the elements that must be retained from those that can vary. It is best to first use a brief prompt to test the direction, and then gradually add more details such as camera angles, materials, etc.; this approach makes it easier to identify issues compared to including a large number of adjectives at once.

Prices, points, and membership

The web version of JoyAI uses a points-based system, and it features two membership levels: regular and premium. On the public page, which requires login to access, the pricing for memberships, the duration of membership periods, the points allocated each month, the concurrent usage capacity, and the conditions for refunds are not displayed consistently.

The free points available each day are not unlimited on a permanent basis; any points that are not used on a given day are not carried over or replenished. The cost associated with specific tasks is displayed dynamically on the page, and differences in models, patterns, and output parameters can result in varying consumption levels.

Notes on Purchases and Refunds

• Confirm the package name and membership level before making the payment.
• Check the one-time purchase or automatic renewal cycle.
• Check the points earned, their validity period, and the rules for resetting them.
• Verify the actual points for image, video, and digital human tasks.
• Save the payment page and order records.
• The methods for cancellation and eligibility for refunds are subject to the rules of the payment channel.
• The public page does not specify a deadline for refunds; therefore, it should not be assumed that refunds can be granted unconditionally.

Differences between the web version and the iPhone app

The iPhone app is currently provided by Suqian Bànxīn Intelligent Technology Co., Ltd.; it requires iOS 15 or a higher version. The app is available in both Simplified Chinese and English in the app store, and it is labeled as suitable for users aged 18 and above. The web version does not specify identical compatibility requirements or age restrictions, so those specifications cannot be applied directly.

API, SDK, GitHub, and open-source status

Jingdong making the model code and weights available publicly is a resource intended for independent developers; it is not possible to determine the specific model used on a website based on this information, nor can the license associated with those repositories be extended to the content generated by the website. When deploying something on one’s own, it is necessary to read the licenses of the target repository, the model weights, and any dependent components.

Privacy, security, and copyright

The website currently provides links to the privacy policy, user agreement, and community guidelines; however, on the public page without logging in, information regarding data retention periods, uses for training, deletion procedures, and corporate data policies is not displayed in its entirety. Sensitive materials should be submitted only to the extent necessary.

• Do not upload identification documents, bank cards, medical records, or confidential business information.
• Appropriate licenses are required for real-person portraits, voices, and voiceover scripts.
• Third-party images, fonts, trademarks, and characters may be protected by copyright or brand rights.
• The fact that input materials are available does not mean that the resulting output automatically comes with commercial licensing.
• Do not remove the AI-generated identifiers required by law.
• Check the characters, facts, text, and potential copyright infringement elements before downloading.
• For important projects, input authorization and manual review records should be retained.
• The actual effects of account logout or work deletion are subject to the current terms and conditions as well as the backend system.

Boundaries for commercial use

The public pages do not provide a unified commercial licensing scheme that can cover all models, materials, and membership levels. For advertising, e-commerce, publishing, and client projects, it is necessary to verify the benefits of the current package, specify the licensing terms for the materials used, define the relevant terms of use, and obtain authorization for any individuals involved before proceeding with the publication.

Model generation is probabilistic, and it may produce elements similar to existing works or incorrect brand information. Before use in commercial purposes, manual copyright checks are required, and the outputs should not be presented as verified facts or actual product images.

Suitable for users and scenarios

• Content creator: Creates illustrations, video clips, and draft audio recordings.
• Marketer: Explore directions for visual and social content related to campaigns.
• E-commerce team: Generates background and conceptual materials that require manual verification.
• Designer: Use zooming, redrawing, and image masking to speed up draft iterations.
• Short-video team: Creates storyboard tests and dynamic creative prototypes.
• Individual users: Learn about AIGC workflows using daily points.
• It is not suitable for directly handling highly confidential information, unauthorized portraits, or content that requires absolute accuracy.

Product advantages

• Tasks involving covering images, videos, and digital avatars on a webpage.
• The generated assets can be further edited and managed in a unified manner.
• It offers options for redrawing, removing elements, image masking, enlarging images, and achieving high-definition quality.
• It supports various video input methods.
• Daily logins earn points that lower the threshold for accessing the lightweight trial version.
• You can directly copy the image or download the generated file.
• The canvas is suitable for organizing multiple rounds of visual exploration.

Main limitations

• Free points are reset on the day they are earned and cannot be accumulated.
• The points for each task are dynamically configured on the page.
• The public selling price for members and the full range of benefits have not yet been determined.
• The canvas is still marked as Beta.
• The generated results may contain factual, textual, and structural errors.
• Video character consistency and complex movements cannot be guaranteed.
• The use of digital humans involves obligations regarding portraits, voices, and identifiers.
• The web version does not support public APIs and SDKs.
• Open-source model projects are not the same as open-source web versions.
• Insufficient details are provided regarding data retention, training purposes, and deletion.
• A commercial license cannot be inferred merely from something that can be downloaded.

Frequently Asked Questions

What is the web version of JoyAI?

It is an AIGC creative workshop operated by the JD.com ecosystem, used primarily for creating images, videos, and digital avatar content, as well as providing tools for canvas management and asset handling.

Is the web version of JoyAI owned by JD.com?

Yes, the resources on this page, the login system, the service interfaces, and the registration information all point to the JD.com system.

Is the web version of JoyAI free?

It is possible to experience the service on a limited basis by earning points through daily logins, but this is not unlimited free access forever; the current guidance indicates that 68 points are awarded each day and they are reset at the end of that day.

Are the 68 points earned each day accumulated?

No, the free points available for that day are automatically reset at 24:00; they are not accumulated, carried over, or reissued.

How many points are required to generate one time?

The consumption for different models, tasks, and parameters is displayed dynamically on the creation page; it is necessary to check this consumption before submitting.

How much is the membership fee?

On the public, unlogged-in page, the prices for regular members and premium members are not displayed consistently; it is necessary to refer to the member account center and the settlement page for accurate information.

Which image formats are supported?

Currently, the upload verification supports JPG, JPEG, and PNG formats; each image must not exceed 10MB in size.

Can videos be created?

Yes, the web page can contain text, images, beginning and ending frames, as well as video creation tasks that involve multiple images; the specific format follows that of the current creation page.

Can digital voice broadcasting be done?

Yes, users can enter the text that the character is supposed to say and generate a video, but it is necessary to obtain authorization for the use of the portrait, voice, and script.

Can the generated content be used for commercial purposes?

The public page does not provide a unified commercial license that covers all models and packages; it is necessary to check the current terms, material licensing rules, and settlement provisions before publishing.

Is the web version of JoyAI open source?

The source code of this web-based product is not made public; JD.com also has open-source projects such as JoyAI-Image, and these two should not be confused with each other.

Are there public APIs or SDKs available?

To date, there are no confirmed web-based public APIs for ordinary developers, no process for applying for keys, and no official SDKs available.

Is the web version the same as the iPhone app?

It’s different: the web version focuses on AIGC-based creation, while the iPhone app offers a range of functions such as conversation, news, shopping, and companionship.

How to add a FAQ Schema

Visible FAQs are provided on this page; structured FAQ data should be generated uniformly by the website template or the backend, and script tags cannot be embedded in the text.

Summary

The web version of JoyAI combines images, videos, digital avatars, and generative editing tools within a single browser interface, making it suitable for quickly exploring different content options and further processing assets. 68 points are available each day to enable a basic usage experience, though these points are reset at the end of each day.

Before deciding to use it on a long-term basis, one should check the actual price, the amount of resources required for tasks, options for refunds, and any commercial benefits in the member area after logging in. When dealing with real people, customer materials, or content that is to be made public, it is also necessary to address issues related to privacy, portraits, copyrights, and AI identifiers.