What is Cicada Mirror AI?
Chanjing AI is an AI digital human and short-video creation platform developed by Xiamen Chanjing Technology Co., Ltd.
It covers imagery, audio, copywriting, voiceovers, assets, subtitles, batch production, and API integration.
Main functions
- Choose to generate spoken video using a public digital avatar.
- Clone a digital avatar using real-person video.
- Upload an image to create a digital avatar based on a photo.
- Generate digital human avatars through textual descriptions.
- Clone sounds or create custom timbres.
- Convert PPT or PDF files into explanatory videos.
- Create lightweight voiceovers using photos to convey messages.
- Generate products, knowledge, and matrix videos in bulk.
- Add subtitles, backgrounds, music, pictures, and videos.
- Access digital human customization and synthesis through APIs.
Types of digital humans
| Type | Creation method | Suitable scenarios |
|---|---|---|
| Public digital humans | Simply select the existing image of the platform. | Quick trial and general narration |
| Cloning of real humans into digital versions | Upload real-person videos that meet the requirements. | Brand founders and instructors |
| Digital image avatar | Upload a single image of a person | Lightweight content and virtual avatars |
| Text-to-digit human avatars | Text descriptions generate images. | Character creativity and marketing materials |
| Highly precise customized digital humans | Contact for business data collection and production | Brands, companies, and long-term projects |
The realism, range of motion, background capabilities, and price of real people, photos, and AI-generated images vary; they should be tested on a case-by-case basis.
Digital voice broadcasting
Digital voiceover involves using text or audio to have characters give explanations, with subtitles and other elements added in an editor.
| Stage | Main operations | Key points of inspection |
|---|---|---|
| Copywriting | Enter text or upload audio | Facts, pronunciation, and duration |
| Characters | Choose a public or custom digital avatar. | Authorization and background capabilities |
| Sound | Select a timbre or clone a sound | Right to sound and pronunciation |
| Image | Set proportions, background, and materials | Composition and brand guidelines |
| Subtitles | Automatically generate and manually proofread | Proper nouns and punctuation |
| Export | Select the clarity level and generate the video. | Watermarks, resolution, and cicada beans |
Converting PPT and PDF to video
The platform can convert presentation or document pages into videos narrated by digital avatars, making them suitable for courses, training sessions, and product introductions.
- Clear hidden pages and irrelevant notes before uploading.
- Standardize fonts, proportions, images, and copyrights.
- Break long pages into sections suitable for oral delivery.
- Add natural explanatory text for each page.
- Check page transitions and speech rhythm.
- Check the subtitles and media content page by page before exporting.
Voice cloning
Voice cloning creates vocal timbres based on authorized recordings; it can be used to drive videos as well as for standalone voice synthesis.
- Only clone one’s own voice or obtain verifiable authorization.
- Record samples that are quiet, clear, and free of reverb.
- Avoid uploading customer calls and private recordings.
- Test numbers, English, place names, and technical terms.
- The public content should indicate that synthetic sound is being used.
- Set access permissions for timbres and define the deletion process.
Chanjing AI Usage Guide
- Register and log in to see the 60 free Cicada Beans and a 3-day period.
- Choose digital voice narration, PPT to video conversion, or voice synthesis from photos.
- Upload materials that have been authorized and have had their sensitive data removed.
- Choose a public digital avatar or create a custom one.
- Enter text or upload audio and set the sound.
- Verify the background, proportions, subtitles, and visual elements.
- View the estimated consumption of chickpeas that is displayed before submission.
- Generate short videos to test lip movements, actions, and pronunciation.
- Correct proper nouns, pauses, and subtitles.
- Batch tasks should be created only when mass production is required.
- Select the clarity level and complete the final export.
- Save authorizations, scripts, task numbers, and the final output.
Free version and quota for new users
| Project | Current instructions | Precautions |
|---|---|---|
| Registration gift | 60 Cicada Beans | Valid for 3 days |
| Digital human | Some public or customized features are available for experience. | The number of times and the images available are limited. |
| Export | It may have a brand watermark. | To remove watermarks, the paid version is required. |
| API | The version list indicates support. | The actual key, quota, and concurrency levels need to be confirmed. |
| 4K | Not supported | Available in Professional version and above |
| Batch creation | Not supported or limited | High-frequency production requires an upgrade. |
The 60 Cicada Beans are suitable only for short tests; it is recommended to first check the appearance, voice, and lip movements before deciding on a package.
Price of the basic version
The prices listed below were verified as of August 31, 2026; continuous subscriptions and one-time purchases need to be compared separately.
| Purchase plan | Price | Total cicada beans | Reference video duration |
|---|---|---|---|
| Trial version for 1 month | 48 yuan | 600 | About 10 minutes |
| Trial continuous monthly subscription | 30 yuan | 600 | About 10 minutes |
| Basic version: 1 month | 98 yuan | 2000 | About 33 minutes |
| Basic continuous monthly subscription | 78 yuan | 2000 | About 33 minutes |
| Basic version: 6 months | 468 yuan | 12000 | About 200 minutes |
| Six months of consecutive subscriptions | 438 yuan | 12000 | About 200 minutes |
The basic version is suitable for individual creators and small businesses; it offers features such as watermark removal, the ability to create custom digital avatars, and limited batch processing capabilities.
Professional version price
| Purchase plan | Price | Total cicada beans | Reference video duration |
|---|---|---|---|
| 1 month | 758 yuan | 20000 | About 333 minutes |
| 3 months | 2168 yuan | 60000 | About 1000 minutes |
| 6 months | 4188 yuan | 120000 | About 2000 minutes |
| 12 months | 9,998 yuan | 300000 | About 5,000 minutes |
The Professional version is aimed at those who create content frequently, as well as at account managers and developers; its key features are 4K quality, concurrent processing capabilities, limits on the number of tasks that can be processed simultaneously, and fast generation speeds.
Enterprise version price
| Purchase plan | Price | Total cicada beans | Method of distribution | Reference duration |
|---|---|---|---|---|
| Enterprise version: 12 months | 32,800 yuan | 1200000 | 100,000 beans per month | About 20,000 minutes |
The Enterprise version is suitable for team asset management, large-scale production, and higher API concurrency; the contract should also specify the number of seats, SLAs, and services.
Comparison of version benefits
| Equity | Free version | Basic version | Professional Edition | Enterprise Edition |
|---|---|---|---|---|
| Download watermark | With brand watermark | Removal by default | Removal by default | Removal by default |
| Custom digital humans | Limited experience | Support | Support | Support |
| 4K video | Not supported | Not supported | Support | Support |
| Batch creation | Not supported or limited | Small batches | Larger batches | Team-level batch processing |
| API | Support | Support | Support | Support |
| Team management | Not supported | Not supported | Not supported | Support |
The reference duration is not a fixed output amount.
The official form estimates that 1 minute of video can be produced from 60 cicada beans, but the actual time required may be higher or lower.
- The pricing for different types of digital humans may vary.
- 4K and high-resolution modes may increase consumption.
- Additional fees may apply for audio, storyboard materials, and AI features.
- Each video in a batch task is billed separately.
- Before submitting a task, refer to the estimated time required on the page.
- In the case of a failure and a return, it is necessary to check the details of the Cicada Beans.
Distribution and validity period of Cicada Beans
| Source | Issuance and Validity Period | Impact on equity |
|---|---|---|
| Gift for new users | 60 beans, valid for 3 days | For trial use only |
| Member packages | Paid on a monthly basis, valid for the current month | Benefits are provided depending on the membership level. |
| Add Dou Bao | The information indicated on the purchase page shall prevail. | Only add cicada beans, without upgrading the version. |
| Business trial | In accordance with the business agreement | Used for API or enterprise testing |
| Large-scale purchases | Subject to the validity period of the contract. | Can be integrated with corporate solutions |
The total amount for the package may not be credited all at once; any unused balance that remains across months might become invalid, so it is necessary to check the date of issuance before making a purchase.
API capabilities
Member users can obtain API keys and documentation in their personal account; the billing is based primarily on the number of seconds of video generated.
| API capabilities | Explanation |
|---|---|
| Customized real-person digital avatars | Upload a video to create an avatar. |
| Voice cloning | Create a custom playable tone. |
| Video synthesis | Text or audio-driven digital humans |
| Public models | Select the platform for generating videos with digital avatars. |
| Audio synthesis | Separate charging when no video is generated |
| Callbacks and tasks | Submit according to the document and receive the results. |
Video synthesis usually already includes voice synthesis; when only sound generation is requested, resources are consumed in accordance with the relevant rules.
API Integration Tutorial
- Select members who meet the requirements in terms of clarity, concurrency, and team needs.
- Obtain the key and the current API documentation in the personal center.
- Store the key on the server side, without including it in the frontend pages.
- First, create or select an avatar and voice that can be used.
- Submit short text synthesis tasks in the testing environment.
- Save the task ID and verify the callback correctly.
- Check the actual number of seconds, as well as cases of failure and subsequent rejections.
- Add timeout, retry, idempotency, and rate limiting mechanisms.
- Obtain permissions for the image, voice, and copy before going live.
- Continuously monitor concurrency, costs, and content security.
Supported platforms
| Platform | Current status | Requirements |
|---|---|---|
| Web | Complete creation and open platform | Modern browsers |
| Mini-program | It is possible to view and handle certain tasks. | Based on the current entry point. |
| iPhone | Official App | iOS 15 or higher |
| iPad | Official App | iPadOS 15 or higher |
| Apple Silicon Mac | The iOS version is available for use. | macOS 13 and above, M1 and later |
| Apple Vision | Compatibility | visionOS 1.0 or higher |
Prices may vary on websites, the App Store, and other payment platforms; therefore, it is not possible to use the amounts from different packages together.
Authorization and Compliance
- To clone a real person, it is necessary to obtain permission for their portrait, voice, and personal information.
- Employees, customers, and actors should be made aware of the scope of use and the duration.
- It is not allowed to use digital avatars to pretend to be real people in order to carry out deception.
- The generated content should indicate its AI origin in accordance with the applicable rules.
- Images, music, fonts, and text must have legitimate sources.
- Professional compliance audits are conducted in cases involving advertising, healthcare, and finance.
- After the authorization is revoked, new asset creation should be halted, and existing assets should no longer be processed.
The boundary between APIs and open source
| Project | Status |
|---|---|
| Open API | Provided by the authorities; members can obtain keys and documents. |
| Web page and API features | Differences may exist; it is necessary to check according to the document. |
| Official SDK | No public, general-purpose SDK repositories were found. |
| Official GitHub | No verifiable open-source product organizations were found. |
| Platform and model source code | Not disclosed; the availability of an API does not mean that the product is open source. |
Usage restrictions
- The lip movements, actions, pronunciation, and subtitles may be incorrect.
- The reference minutes do not guarantee the amount of output.
- The generation speed and concurrency may vary during peak times.
- The package with cicada beans is valid for the current month; it may become invalid if not used.
- \"Unlimited\" or exempted benefits may be subject to fair use rules.
- Digital avatars cannot replace genuine endorsement and professional review.
Frequently Asked Questions
What kind of videos is Cicada Mirror AI suitable for creating?
Suitable for digital voiceovers, course training, product explanations, knowledge sharing, PPT presentations, and matrix short videos.
Do new users have a free quota?
Yes, a registration currently comes with 60 Cicada Beans as a gift, valid for 3 days; this is suitable for gentlemen who want to test out short videos.
How much is the basic version of Chánjìng AI?
The basic subscription starts at 30 yuan per month, while the full basic version costs 78 yuan per month; the specific amounts and renewal details are indicated on the settlement page.
Are the reference minute counts in the package fixed?
No, the actual consumption is influenced by the model, clarity, type of digital human, and the AI functions used.
Can the cicada beans be carried over to the next month?
The membership packages with cicada beans are delivered on a monthly basis and are valid for that month; for beans from other sources such as BeanBao, the details are indicated on the purchase page.
Does Chianjing AI provide APIs?
Members can obtain the keys and relevant documents in their personal account; the service supports customizing digital avatars, voice cloning, and video synthesis.
Is Chanting AI an open-source product?
No, having an open API does not mean that the platform or the models are open source; to date, no official open-source product repository has been identified.
Guigong Network Security Registration No. 45132202000164