Dream Creation Canvas

What is the Dream Creation Canvas?

DreamCreate Canvas is a platform for generating AI images, videos, scripts, and music; its core feature is an infinite canvas on which nodes can be placed and connected freely.

Users can combine prompts, generation models, parameters, reference materials, and result nodes to create reusable workflows, which are suitable for expanding from a single image to storyboards, short videos, and bulk visual content creation.

Current product structure

Infinity canvas and node workflow

Node orchestration

• The canvas can hold elements such as text, settings for image generation, settings for video generation, the resulting images and videos, as well as storyboard plans.
• Users determine the data flow by establishing connections; for example, they send text prompts to the image generation node, and then send the resulting images to the video generation from image node.
• The node locations and parameters are saved along with the project, making it easy to return to the same process for further adjustments.

Intelligent workflow planning

• The AI architecture within the canvas allows for the design of node diagrams based on specific requirements, and it can determine which approach is suitable for tasks such as text-to-image generation, text-to-video generation, storyboarding, or converting scripts into videos.
• For longer stories, intelligent storyboard planning tools can be used to break the text down into multiple scenes, after which images and videos can be combined to create the final content.
• Automatic planning is just a starting point; the user still needs to check the node types, connection directions, parameters, and the correspondence between the various elements.

Templates and reuse

• Workflows can be saved as templates, preserving nodes, connections, positions, and configurations to avoid repeated setup.
• Templates are suitable for repetitive tasks such as brand visual design, fixed character usage, batch image creation for products, and series of short dramas.
• When reusing it, it is necessary to verify again whether the model is still usable, as well as whether there have been any changes in the integration rules, dimensions, and output parameters.

Complete steps for setting up a workflow

1. Register an account and create a canvas project, first deciding whether you need images, videos, or a complete storyboard.
2. Add a text node and specify the subject, setting, style, composition, camera angle, and constraints.
3. Add image or video configuration nodes, and select the model, aspect ratio, clarity, duration, and number of batches.
4. Upload the reference image and connect it to the corresponding generation node to verify the reference relationship between roles, products, or scenarios.
5. To connect the result nodes, first run a small number of low-cost sample checks.
6. Adjust the prompts, models, and parameters; once the results are stable, improve the image quality or proceed with batch processing.
7. Save the mature processes as templates, and archive the models, parameters, materials used, as well as the time of creation.

AI image generation and editing

Text-to-image and reference images

• The user enters prompts in Chinese or in other languages supported by the model, selects the model and image quality, and then the image is generated.
• Reference images can be used to maintain consistency in the appearance of characters, details of products, the environment of a scene, composition, or style.
• Different models have varying capabilities in text rendering, realism, animation, local editing, and maintaining consistency with references.

Current image model

The list of models and their pricing are the data displayed on this page; the supplier, version, cost, and image quality settings can change at any time.

Consistency between characters and scenarios

• Character design allows for the organization of a character’s appearance, clothing, posture, and various perspective views into materials to be used in subsequent shots.
• Scene generation is used to create the environment, color palette, lighting, and camera scale, which can be referenced repeatedly in the storyboard.
• Reference consistency is not an absolute guarantee; the hand movements, text, details, and the appearance across different shots still require manual review.

AI video generation

Current video models and scores

• The currently available models include Seedance 2.0 and Fast, HappyHorse 1.0, Vidu Q3, and the Grok video series.
• The basic consumption is calculated on a per-second basis; currently, it is around 3 to 10 points per second.
• 480P, 720P, and 1080P may use different scaling factors; for some 1080P content, the calculation is based on up to 3 times the standard consumption.
• The final cost is also influenced by the video length, the number of batches, the model used, and the quality of the video; it is necessary to check the estimated charges before submitting.

Short plays, scripts, and storyboards

• The short drama module includes separate pages such as a project list, the project homepage, assets, videos, and an editor.
• Users can enter a story or script, and the AI will analyze the characters, conflicts, scenes, shots, and visual rhythm.
• The basic cost of using the AI storyboard planning model is 1 point per use; subsequent image and video generation are charged according to their respective models.
• Character assets and scene assets can be reused across different shots, reducing the effort required to organize continuous content.
• The script structure, dialogue, duration, continuity of shots, and narrative pace still require manual review by the screenwriter and director.

Short drama workflow

1. Create a short drama project by entering the theme, target audience, number of episodes, duration of each episode, and the core conflict.
2. Create or import a script to identify character relationships, scenes, and key plot points.
3. Execute the storyboard planning by breaking the text down into shots, shot sizes, actions, and visual cues.
4. Create role and scene assets to provide consistent reference images for the shots.
5. Generate storyboard images in bulk, eliminating those frames that fail the criteria regarding appearance, composition, or continuity.
6. Send the selected storyboard to the video node, and set the motion, duration, scale, and image quality.
7. Add music, sound effects, dialogue, and editing, then export the final video.

Music, sound effects, and voices

E-commerce and commercial visual design

• The mobile and web versions of the code contain modules for product images, e-commerce visuals, logos, posters, and materials suitable for various platforms.
• The batch generation feature for product images allows the same product to be adapted with different backgrounds, proportions, and marketing scenarios.
• The multi-platform module allows materials to be organized according to the specifications of different channels, but text, safety zones, and advertising rules still need to be checked separately for each platform.
• Portrait retouching is used to improve the visual appearance of a person; it cannot be employed to simulate a realistic image or to create portraits without consent.
• The sprite extraction tool can isolate elements from composite images; it is necessary to verify the rights to the original image before using it.

Inspiration Assistant and Prompt Words

• Inspiration generation, creative coaching, and aesthetic testing are currently charged at 1, 2, and 1 points per use, respectively.
• The platform includes built-in personalized assistants for visual creativity, scriptwriting, short videos, and commercial advertising.
• Assistants can transform vague requirements into prompts for composition, lighting, style, narrative, and modeling.
• The suggested prompts may contain a style or marketing approach that is not suitable for the project; they should be modified in accordance with the brand guidelines before use.

Subscription and monthly card prices

The interface fields for the premium version indicate 3,600 points and 50GB of storage per month, while the description of the benefits states 3,300 points and 100GB of storage per month; there is a discrepancy between these two figures.

Before purchasing the premium version, it is necessary to obtain written confirmation from the payment page, the account statement, and customer service regarding the points and storage available; one should not rely solely on estimates based on a single set of figures.

Points, sign-in, and refunds

• Different models for images, videos, conversations, music, and short dramas require different amounts of points; higher resolution content and bulk processing may involve multiplier factors.
• The 600 points associated with the monthly card are credited immediately, while the additional 600 points must be obtained by logging in daily over a period of 30 days.
• 20 points are awarded each day; any points not earned on a given day are lost, and it cannot be assumed that the maximum of 1200 points will be credited automatically.
• The reward points purchased at one time are valid forever, and some recharge packages may include additional points as a gift.
• Recharge points are virtual goods, and no refunds are available once they have been purchased; in the event of abnormal deductions due to platform failures, the amount can be refunded after verification.
• Generating content in violation of the rules may result in the loss of points or the suspension of the account, with no refunds provided.

Custom APIs and supplier configurations

Multiple API configuration sets

• Advanced settings allow the configuration of multiple sets of external AI interfaces, which can be designated for use in conversations, with images, and for videos.
• Users can enter the supplier, service address, and API Key, and set a primary source and a backup source.
• If the primary source fails, it is possible to switch to the next set of configurations, which is suitable for workflows that require fault tolerance.

Process of using your own key

1. Create a separate project and a restricted key at the model provider.
2. Set budgets, call limits, allowed models, and risk alerts.
3. Add suppliers, service addresses, and keys to the DreamCanvas.
4. For each group, select the purpose for the dialog, image, or video, and set a priority level.
5. Use a small number of samples to test authentication, model names, timeouts, retries, and billing.
6. Check the actual usage amount in the supplier’s backend; if anything is abnormal, revoke the key immediately and replace it.

Data and cost boundaries

• When using external interfaces, prompts, images, or videos may be sent to the selected supplier, whose data policies and regulations apply separately.
• The supplier fees resulting from the use of own keys and the DreamCanvas points may be calculated separately; it is necessary to verify whether double charging occurs before submission.
• Do not save keys with high permissions on shared devices, nor use the unlimited production keys associated with the root account.
• Clearing the canvas settings does not necessarily revoke the supplier key; it is still necessary to go to the supplier’s backend to cancel it after leaving the project.

Platform APIs, SDKs, and open source

• The built-in third-party API configuration is a feature that enables calls from the client side; it does not mean that DreamCanvas provides public developer APIs.
• At present, the API documentation, key application procedures, pricing, or service levels for DreamCanvas intended for external developers have not been confirmed.
• Currently, no official SDK, plugin marketplace, or public example projects have been confirmed.
• It has not yet been confirmed whether the DreamCanvas website, Infinite Canvas, and the generation services use open-source licenses.
• Other canvas projects with the same name or underlying open-source components do not constitute proof that this product as a whole is open source.

Materials, works, and storage

• The asset library can store uploaded files, as well as images and videos created recently, along with relevant prompt texts, model information, and size details.
• The project page is used to view, download, and manage the generated results; short-form video projects also include characters, scenes, and video assets.
• The subscription plans offer different storage capacities; additional expansion may incur extra fees in accordance with the rules valid for the current month.
• Important projects should be saved locally or in the company’s asset repository as well; online accounts should not be used as the only form of backup.
• The data processing rules for deleting items, materials, handling expired subscriptions, and situations of storage limit exceeded should be confirmed with customer service before carrying out such operations.

Copyright, Commercial Use, and Generation Identifiers

• The top-up agreement states that the copyright to the images and videos created belongs to the user, but the user must comply with the law and ensure that the materials used are legal.
• The premium version explicitly lists commercial licensing options; however, there is no clear public indication as to whether the free, standard, advanced versions, and the option based on usage points also cover commercial use.
• When a commercial release is required, the current package licensing terms, the specific model conditions, and the rights regarding the materials shall apply.
• The output of the model may resemble existing works, characters, brands, or product designs; the platform’s copyright information does not rule out the rights of third parties.
• The home page indicates that the content on the platform is generated by AI; nevertheless, users must still comply with the relevant requirements regarding the labeling of AI-generated content when posting it externally.
• It is prohibited to generate content related to politics, pornography, violence, infringement of rights, or any other illegal or unlawful material.

Privacy and security considerations

• Registration accepts usernames, phone numbers, verification codes, and passwords; logging in can also be done using a phone number or email address.
• The canvas handles prompts, the uploading of images and videos, the project structure, the generated results, and model configurations.
• The external APIs configured by users, along with the credentials for object storage, are highly sensitive information; therefore, the principle of least privilege and the use of separate keys should be applied.
• At present, there is no publicly available, confirmed version of the full privacy policy, information on the duration for which data is stored, details regarding its use in training purposes, rules concerning the timing of data deletion, and guidelines for cross-border data processing.
• When dealing with customer products, unreleased characters, real-person portraits, or trade secrets, it is necessary to first consult the platform regarding the mechanisms for maintaining confidentiality and for deleting such information.
• Payments and transaction records should be retained to facilitate the verification of abnormal charges, authorizations, and project versions.

Which users are it suitable for

• AI creators who wish to manage complex image and video processes using a node-based approach.
• A short-form drama team that requires characters, scenes, storyboards, and the continuous production of videos.
• E-commerce operators who create product images, posters, logos, and materials for various platforms in bulk.
• Design and content teams that need to save mature processes for reuse.
• Advanced users who possess third-party model accounts and wish to consolidate the management of interfaces from multiple suppliers.

Product advantages

• The node canvas clearly shows the dependencies among prompts, materials, models, and results.
• Images, videos, scripts, music, and e-commerce visuals are all integrated within the same project framework.
• Reusing templates and assets helps maintain consistency in characters, scenes, and brand visuals.
• Subscriptions, monthly cards, and permanent points are all available, allowing users to choose the option that best suits their level of activity.
• It allows the configuration of multiple sets of external interfaces and backup sources, making it suitable for users who have their own model resources.

Capabilities boundaries and limitations

• Complex workflows enhance control but also increase the costs associated with nodes, parameters, integration, and error resolution.
• Changes in models, integrals, magnifications, and availability may prevent old templates from being reproduced at their original cost.
• Consistency of characters and products across different shots may still vary, so manual screening and image editing are essential.
• Video generation is calculated on a per-second basis, with a clarity multiplier applied; high image quality and batch processing result in higher costs.
• There are conflicts in the public fields for the premium version; it is necessary to verify the actual amount received before making the purchase.
• General privacy, refund, and terms related to lower-tier packages have limited transparency, so they are not suitable for use in projects that require high levels of confidentiality.

Frequently Asked Questions

Is Mengchuang Canvas an ordinary AI image generation tool?

Not only that, but it is based on a node-based infinite canvas, and it also covers video, script storyboarding, music, e-commerce visuals, as well as the management of assets and templates.

Can DreamCanvas generate videos?

Yes, it supports text-to-video generation, the first frame, the first and last frames, as well as the reference material mode; the specific options depend on the capabilities of the chosen model.

How much is the DreamCreate Canvas?

The current standard monthly plan starts at 29.9 yuan; there is also a monthly card option that costs 66 yuan, as well as the possibility to top up credits as needed. The prices for annual plans and higher-tier packages can be found on the purchase page.

Do the points expire?

The points acquired through a single top-up are valid indefinitely, but the points earned by signing in daily with a monthly subscription become invalid if not claimed on the same day.

Can a refund be obtained after topping up with points?

No, the recharge terms specify that virtual points purchased cannot be refunded; any deductions due to platform issues can be reversed once verified.

Can the content generated by DreamCreate Canvas be used for commercial purposes?

The premium version includes a commercial license; the details for other packages are not fully specified. For commercial projects, it is necessary to obtain written confirmation of the appropriate package and model before making a purchase.

Can I use my own API Key?

Multiple sets of external model interfaces can be configured, each dedicated to conversations, images, or videos, with the option to switch between primary and backup versions.

Does Mengchuang Canvas offer a public API?

It is currently confirmed that users configure third-party APIs; it has not yet been determined whether DreamCanvas offers public interfaces, SDKs, and service levels for developers.

Is DreamCreate Canvas an open-source project?

At present, it has not been confirmed that the core products are open source; moreover, it is not possible to determine whether the entire platform is open source by relying on the licenses of other projects with the same name or of the underlying components.

How to add a FAQ Schema

The common questions on this page are presented using hierarchical headings, with the answers visible right next to the corresponding paragraphs; the website can use this information to generate structured FAQ data on the server side.

Structured data should be consistent with the page content; hidden prices, fixed lists of models, commercial commitments related to lower-tier packages, public APIs, or undisclosed data policies should not be included.

Summary

DreamCreate Canvas enables the transformation of images, videos, storyboards, music, and commercial visuals from being created once-only into visual, reusable node-based workflows.

Before going into full production, it is necessary to first test the model and processes using low-cost samples, and to verify the integration factors, commercial licensing, API keys, as well as the boundaries of the stored data.