Decohere AI
Decohere AI – makes AI programming more efficient and simpler.
Tags:AI programming toolsWhat is Decohere AI?
Decohere AI is a web-based generative visual tool developed by Decoherence, Co.; it is used to create images, characters, and short videos based on text or reference images. It features real-time preview and continuous changes, making it suitable for creators who need to quickly try different styles, select visuals, and proceed with video production.
The product is not currently an AI programming platform, nor does it offer any APIs. Users can register via a browser; they can use the free real-time image generation feature, or they can purchase credits to unlock additional functions such as video processing, zooming, watermark removal, and character training.
Main functions
- Real-time AI imaging: By entering text prompts, you can watch the images change over time; you can modify the descriptions while simultaneously filtering the composition, characters, and style – making it ideal for exploring ideas and carrying out rapid iterations.
- Quality images and templates: Grouped images are generated using prompts or predefined templates; generating one quality image or template costs 1 point, and 4 images are returned as a result.
- Tusheng Video: First, create a starting image on the platform or upload your own image, and then control the direction of movement to produce a short video; each video requires 1 point, and the length of the video is 4 seconds.
- Creative Zoom: enlarges images or videos by 4 times and adds details through an adjustable level of detail enhancement; 2 credits are spent each time this operation is performed, and the result may alter the texture or certain aspects of the original image.
- Reference Person: After uploading a photo of a person, the characteristics of that person are retained in the new prompts and scenarios; it is suitable for use as a profile picture, for creating virtual characters, and for producing sequential content.
- Custom AI characters: Design the body, select a face, and generate the character image step by step; creating a custom character requires 50 points, while generating a set of 4 character images costs 2 points.
- AI Stations: Allows the creation of a continuous stream of generated images based on specified pictures or styles, enabling the exploration of numerous variations in similar styles; it is useful for discovering suitable color schemes, compositions, and visual elements.
Input and output
| Task | Main inputs | Output | Public consumption already disclosed |
|---|---|---|---|
| Real-time images | Text prompts, style direction | Continuously changing AI images | The free version has a daily limit; the paid version offers unlimited real-time images. |
| Quality image or template | Prompt or template parameters | 4 images per time | 1 point |
| Character image | Character design, faces, or reference figures | 4 character images per time | 2 points |
| Video generation | Platform-generated images or image uploads, motion control | 4-second videos | 1 point |
| Zooming in on images or videos | Existing images or videos | 4x resolution media | 2 points |
| Custom role training | Character body and face settings | Characters that can be generated repeatedly | 50 points |
The specific base models, upload formats and size limits, fixed output resolution, and list of prompt languages have not yet been made public. Before uploading, it is necessary to check the file restrictions on the account interface, and use a small amount of credits to test the stability of characters, text, hands, and complex movements.
Usage tutorial
- Log in using your email address, Google, or Apple account. First, enter a brief prompt in the real-time generation area, observe the changes that occur, and then gradually add elements such as the subject, background, camera angle, and style.
- Select images from the real-time results that are close to the desired outcome, then proceed to generate quality maps, apply templates, or use reference figures to maintain the same character.
- When continuous characters are required, create custom characters; be aware that training will cost 50 points, after which multiple sets of poses and scenarios need to be generated to verify consistency.
- When dynamic content is needed, select a platform image or upload your own image, set the direction of movement, and generate a 4-second video.
- Apply a 4x zoom to the selected image or video in order to compare different levels of detail, ensuring that any new details do not deviate from the brand’s elements or the appearance of the characters.
- Before downloading, check the watermark, the commercial usage rights associated with the current package, any permissions related to the individuals depicted, and the copyright status of the materials used; when using them for commercial projects, keep records of the invoices, the benefits included in the package, and the date of creation.
Price packages
| Package or version | Price | Billing cycle | Core benefits or quota | Suitable for users |
|---|---|---|---|---|
| Free | 0 dollars | Permanently free | A limited number of real-time images are available each day, and a wait is required before they can be generated; no videos, zooming functions, and no explicit commercial rights included. | Individuals who experience real-time image generation |
| Explorer | $ | Monthly or annual payment | 600 points per year, unlimited real-time images and videos, zooming, watermark removal, and fast generation | Non-commercial interest creators |
| Creator | $ | Monthly or annual payment | 4800 points per year, role creation, faster generation, optional security filtering, and commercial license | Content creators and business projects |
| Director | The monthly fee is 59 dollars; the limited-time annual payment option shows a cost of 29.99 dollars per month. | Monthly or annual payment | 12,000 points per year, fastest creation, character creation, optional security filtering, and commercial license | High-frequency professional creators |
The annual payment page shows the average monthly price; Director also indicates any time-limited discounts that may apply during a certain period. The currency, taxes, total annual cost, and eligibility for promotions are subject to the details on the actual settlement page. Package points do not carry over when they expire, and they are replenished at the start of each billing cycle according to the relevant plan.
Once the credits are used up, it is possible to upgrade or purchase additional credits. The credits themselves do not expire, but a valid payment plan must be in place in order to purchase and use them. The subscription can be canceled at any time, with the plan being downgraded at the end of the current cycle; no refunds are provided for paid plans.
Commercial licensing and copyrights
- The current package page indicates that both Creator and Director come with a commercial license; it also states that the Creator can use the images and videos produced without having to include any attribution. Upon upgrading to Creator status, the media created previously will also receive a commercial license.
- The current plan page specifies that both Free and Explorer users do not have the right to create content; therefore, even though Explorer is a paid plan, it should not be used for commercial projects by default.
- The terms of service for 2023 separate the permissions for paid members from those for unpaid members, which is not entirely consistent with the description of Explorer on the current package page; before using it for commercial purposes, it is necessary to follow the latest version indicated on the account profile, the billing page, and in written confirmations.
- Users are not allowed to upload copyrighted works for which the necessary permissions have not been obtained, and the content created may resemble the works or trademarks of others; the platform does not guarantee that the output will be free from infringement.
- Users grant the platform broad permission to process inputs and outputs, allowing it to use those data without obtaining additional approval in order to provide, maintain, and improve its services; choosing to submit content to public libraries or social media platforms further expands the scope of utilization by the platform.
Platform, API, and open-source status
| Project | Current status | Precautions |
|---|---|---|
| Desktop web page | Support | The main entry point for creation, allowing direct operation via the browser. |
| Mobile version | Mobile web shortcut | It can be added to the home screen of Android or iOS, but it is not an official native app. |
| API and SDK | Not available | Automated tools must not be used to bypass the website interface, and the terms also prohibit automated access and generation. |
| Browser extensions | Not confirmed yet | No official Chrome extensions were found. |
| Open-source status | Proprietary services | No official GitHub, SDK, or platform open-source licenses were found. |
A registered account can be used by only one user, and each user is allowed to have just one account. Sharing login credentials among team members, using batch scripts for data extraction, or generating content automatically without permission may result in service restrictions.
Privacy and security considerations
- The service may collect the username, email address, IP address, usage data, Cookies, as well as any text, images, and video inputs.
- Payment information is collected by third-party payment processors; Decohere states that it only stores confirmation of payments, and does not keep the complete card details submitted by users.
- Personal data may be processed outside the user’s region, and the retention period depends on the purposes of the service, legal obligations, the need to resolve disputes, and the requirements of applicable agreements; no fixed deadline for deletion is specified.
- Users can request the deletion of their personal data, but such deletion is irreversible; materials that are still in the process of being created, as well as character profiles and billing records, should be backed up as needed first.
- The security policy only specifies the use of commercially acceptable protection methods; it does not disclose the specific encryption standards, independent security certifications, or the fixed retention period for the generated data.
- Reference figures, profile pictures, and uploaded media may contain personal or sensitive information; it is necessary to obtain the consent of the persons depicted. Do not upload identification documents, photos of minors, or confidential material for which you do not have permission to handle.
Suitable for users and use cases
- Short-video and social media content creators: Generate 4-second clips, visual transition elements, and vertical-format content segments.
- Character and virtual avatar creators: Train custom characters or maintain consistency in identity using a single portrait photo.
- Illustration, concept, and brand teams: explore styles in real time, then refine the candidate visuals using high-quality images and zoom functions.
- Musicians and editors: They create visual sequences that cannot be filmed in reality, but it is necessary to verify the rights related to music, characters, and trademarks on one’s own.
- Interested users: They can generate a solution in real time for free, and then decide whether to pay for videos and enlargements.
Capacity boundaries
- Currently, the duration of videos is fixed at 4 seconds, which is more suitable for individual footage rather than for creating complete videos directly; longer content still requires external editing and audio processing tools.
- The generated results may show issues such as drifting characters, structural errors, corrupted text, or unnatural movements; moreover, even the reference characters and custom characters cannot guarantee perfect consistency in each image.
- Zooming in to add new details does not mean restoring the original information without any loss; logos, product structures, and identities of people need to be compared manually.
- The free version provides a limited number of real-time images per day; services such as videos, zooming, and watermark removal require payment, and the points earned under the regular billing plan expire.
- Lacking APIs, team accounts, or public development documentation, it is not suitable for workflows that require automated bulk generation and system integration.
Frequently Asked Questions
Can Decohere AI be used for free?
The free version can be used for an extended period to generate a limited number of real-time images per day, but advanced features such as videos, high-quality images, zooming, and watermark removal require a paid plan and credits.
How many points does a video consume?
Currently, each video generated costs 1 point, and all generated videos are 4 seconds long; enlarging an image or video requires an additional 2 points.
Can the Explorer solution be used for commercial purposes?
The current package page clearly lists commercial licensing as available for Creator and Director roles, indicating that Explorer does not have the right to create content. If the content is intended for commercial use, it is necessary to upgrade to a plan that includes commercial licensing, with confirmation given again at the time of payment.
Are unused points carried over?
Points from monthly or annual packages are not carried over and expire at the end of the billing period. Additional points purchased separately do not expire, but can only be used during the valid payment plan period.
Does Decohere have a mobile app or API?
There is no official native mobile app, and APIs are not available for now. Users can add the web app to their phone’s home screen, but it remains a mobile web experience.
Can I get a refund if I cancel my subscription?
It can be canceled at any time and used until the end of the current period, after which it will automatically revert to the free version. The fees already paid for the plan are non-refundable.
Guigong Network Security Registration No. 45132202000164