SenseMirage is a free AI painting and image generation platform launched by SenseTime. It provides SenseMirage, model training and community sharing capabilities. It is a zero-threshold AI image and design tool for the public.

Second Painting

Core parameters and statistics

SenseMirage is a free AI painting platform launched by SenseTime. Its labels are very clear - "free" and "zero threshold", and the official website even directly summarizes the selling point with "it's free as long as you have it". It is not a complex tool for professional designers, but an image generation portal for the public.

Projects	Public Information
Official positioning	Free AI painting and image generation platform
Core Competencies	Vincentian diagrams, style models, model training, community sharing
Underlying support	SenseTime large model system
Delivery form	Web online platform
Key selling points	Free, zero threshold
Place of Belonging	China

Free + Zero Threshold: Miaohua regards "free use" as its core difference, lowering the threshold for the public to try AI painting, which is different from most similar platforms that use points to pay.

Trainable own models: In addition to directly publishing pictures, Miahua also supports training style models based on its own materials and sharing them in the community, allowing the platform to move from "using other people's models" to "training your own style."

User and market recognition

Miaohua relies on SenseTime's large-scale model capabilities and has gained a certain amount of attention among domestic drawing tools with its "free AI painting" positioning, especially among early adopters and light creators. Its community sharing mechanism also attracts some users to upload and reuse style models.

The official has not disclosed precise data such as unified registration volume, activity, or number of community models within the verifiable scope of this article. The relevant scale is subject to official real-time disclosures by SenseTime and Miahua. No numerical inferences are made here.

Cost advantage

The biggest cost advantage of Miahua is “free”:

• Free pictures on the C side: The platform focuses on free cultural pictures. Mass users can generate pictures without any obvious payment threshold. This is its most direct attraction compared to the point-based platform.
• Free model training: Users can train style models based on their own materials without paying extra for customization capabilities (specific training quotas and restrictions are subject to the official website).
• Hidden costs: In the free mode, the image output speed, queuing, resolution or concurrency may be limited; for professional users who pursue high frequency, high definition, and stable production capacity, the uncertainty of the free platform itself is a hidden cost.

The truth about free: Miahua's "free" is very friendly to experience and light creation, but free often means that resources are allocated on demand - queuing during peak periods, fluctuations in image output speed, commercial licensing boundaries, etc. all need to be verified. It is suitable for personal creation, but authorization and stability must be confirmed before using it for serious commercial production.

Main functions

• 文生图: Input Chinese prompt words to generate pictures, focusing on simplicity, ease of use, and zero threshold.
• Style Model: Provides a variety of style models, users can choose different styles to quickly produce pictures.
• Model Training: Supports training of personalized style models based on own materials.
• Community Sharing: Users can share and reuse models and works in the community to form a creative ecosystem.

Expert View: The hidden value of Miaohua lies in the line of "training + community". Simply publishing pictures for free is just the entrance. The real stickiness comes from "training your own style model and reusing it in the community" - this turns the one-time act of publishing pictures into a creative asset that can be accumulated and shared, and is also the key to the platform retaining active users.

Model and version evolution

• Initial release stage (~2023-09): Launched with a free AI painting platform, focusing on zero-threshold Vincentian drawings.
• Model training version (~2024): Online model training and community sharing, strengthening customizability and ecology.
• Current online version (~2026-06): Relying on the SenseTime large model to continuously iterate Vincentian diagrams and style capabilities.

The precise version number and release date of the underlying model have not been officially disclosed one by one. The above is a general outline. It is recommended to verify the current capabilities on the real-time page of the official website.

Technical advantages

• SenseTime large model support: Miaohua relies on SenseTime's "Daily New" large model system, which has the technical foundation of first-tier manufacturers in terms of image generation quality and Chinese understanding.
• Trainable and community-based: Open the model training capabilities to ordinary users, and amplify the style library through community sharing, forming a "platform + user co-creation" content ecosystem.
• Zero Threshold Engineering: Lower the threshold of use with free and simple product design, and cooperate with SenseTime’s computing power to support mass-scale concurrent graph issuance.

How to use

Entrance	Applicable objects	How to use
Web page	Public users	After logging in, enter the prompt word and select the style model to generate the image
Model training	Advanced users	Upload materials to train your own style model
Community	Creators	Browse, reuse and share community models and works

Typical usage steps: Enter the Miaohua webpage → enter the Chinese prompt word → select the style/model → generate and download; advanced users can upload materials in the model training entrance, train and publish their own style models.

Product Pricing

Miahua takes free as its core positioning:

• The main Wenshengtu capabilities are provided free of charge and are open to the public with zero threshold for use.
• Model training and community functions are also mainly free, with specific quota, concurrency and commercial licensing restrictions.

Since the free policy and functional boundaries may be adjusted with operation, this article does not fix specific restrictions. It is recommended to refer to the official website, especially to confirm the authorization terms before commercial use.

Application scenarios

• Popular creative taste: Zero-threshold experience in AI painting, generating avatars, illustrations, expressions, etc.
• Lightweight content creation: Social media images produced by self-media and individual users.
• Style Customization: Advanced users can train exclusive style models to form a personalized creative style.
• Community co-creation: Reuse and share style models and participate in the creative ecology.

Applicable people

• AI Painting Beginners/Popular Users: People who are attracted by “free, zero threshold” and want to get started quickly.
• Light Content Creator: Individual users who need to produce social media images for free.
• Style Customization Enthusiasts: Advanced users who are willing to train their own style models and participate in the community.
• Not suitable for the crowd: Professional production teams that have strict requirements on picture production speed, concurrency and commercial authorization. The resource fluctuations and authorization boundaries of the free platform may not meet the needs.

Summary and Outlook

Miaohua's core competitiveness lies in "SenseTime large models + free zero-threshold + training and community": it uses a free strategy to promote AI painting to the public, and builds a creative ecosystem through model training and community sharing. The current limitation is that there is limited official disclosure of precise data and version nodes. The speed, concurrency and commercial authorization boundaries in the free mode need to be verified by users themselves.

It is worth observing the iteration of Miahua in terms of underlying model quality, training capabilities and community activity in the future. For individual users, it is a good choice for low-cost AI painting; for teams with commercial production needs, it is recommended to confirm authorization and stability before deciding whether to include it in the formal workflow.

Related tools: Midjourney, stable-diffusion

Version Info

• Miaohua online version ：The current online version provides Vincentian diagrams, style models, model training and community sharing capabilities, and emphasizes free use. The bottom layer relies on SenseTime's large model and continues to iterate. There is no official precise release date and specific capabilities. （~2026-06）
• Miaohua model training version ：Miahua has launched online model training and community sharing capabilities, allowing users to train style models based on their own materials and share them with the community, strengthening the customizability of the platform. There is no official precise release date yet. （~2024）
• First edition of Miaohua ：SenseMirage is launched as SenseTime’s free AI painting platform, focusing on a zero-threshold SenseMirage experience that “all you need is hands-on, and it’s free.” There is no official precise release date yet. （~2023-09）