Xiaomi’s first fully modal large model debuts: The final puzzle piece in the multi-front “people-car-home” battle?
```
After Xiaomi released its self-developed AI large model Xiaomi MiMo-V2-Flash at the Human-Car-Home Ecosystem Partner Conference in December 2025, Xiaomi once again accelerated its pace.
On March 19, Xiaomi launched its first all-modal foundation model, Xiaomi MiMo-V2-Omni.
MiMo-V2-Omni is designed as an “executor” with cross-modal perception and GUI (Graphical User Interface) operation capabilities, and can seamlessly integrate with various Agent frameworks.
Previously, this model was blind-tested under the codename “Healer Alpha” on the OpenRouter platform, and in various benchmark tests showed performance comparable to or even surpassing leading closed-source models.
Regarding the model's efficient "launch speed," Lei Jun said: “We are relatively low-key in the AI field, but our actual progress may be much faster than what everyone sees. In AI, our R&D and capital investment this year will exceed 16 billion yuan. I believe that as long as we continue to invest, Xiaomi will surely produce an impressive report card in the AI era.”
As the core person in charge of this model, Luo Fuli also stated directly on overseas social platforms: “Before tomorrow, anyone in the MiMo team who has tested dialogues less than 100 times can leave directly. This move worked. Once the team's imagination is ignited by the capabilities of the intelligent agent system, that imagination is instantly transformed into R&D speed.”
Currently, Xiaomi has provided an API pricing of $0.4 / million tokens for input and $2 / million tokens for output (supporting 256K context).
Xiaomi's ambition clearly goes beyond selling API to developers.
This model has already partnered with Kingsoft Office (WPS), exploring scenarios of text generation and structured data processing.
However, from a strategic perspective, the commercial endgame for MiMo-V2-Omni points to Xiaomi's “Human-Car-Home Full Ecosystem.”
In its vision for the future of MiMo-V2-Omni, Xiaomi also stated it will “continue to promote long-cycle agent planning, real-time streaming perception, multi-agent collaboration, and deeper integration with the physical world.”
If this model can be deeply integrated as the underlying “brain” into Xiaomi's HyperOS, truly building an AI foundation that can deeply understand voice commands across platforms, independently call mobile apps, and even control the Xiaomi car interface, it will greatly enhance the premium capability of Xiaomi hardware and user retention rate.
Despite the very attractive technology demonstrations and ecosystem visions, Xiaomi is currently facing severe challenges in resource allocation and cost control.
Currently, Xiaomi is in a high-pressure “multi-front battle” situation:
On one hand, its cash-cow smartphone business is being hit by surging upstream memory chip prices, and its comprehensive hardware gross margin is under pressure; on the other hand, the automotive business is at a critical stage of ramping up production capacity and expanding its nationwide sales network, with a pressing need for continued investment.
Furthermore, compared to pure internet giants with fat profit margins and massive cloud computing foundations, Xiaomi does not have an advantage in capital stakes in the AI arms race.
From a strategic vision perspective, MiMo-V2-Omni is undoubtedly the most crucial piece for Xiaomi to complete the intelligent closed-loop of its “Human-Car-Home Full Ecosystem.”
In the headwinds of memory price hikes, how to balance investments across smartphones, automobiles, and large model foundations is a test of Xiaomi management's wisdom.
Risk warning and disclaimerThe market is risky, and investment needs caution. This article does not constitute personal investment advice and does not take into account the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article fit their particular circumstances. If you invest accordingly, the responsibility is yours. ```