Xiaomi
CoverageRelease Notes
Xiaomi unveiled MiMo-V2-Pro, MiMo-V2-Omni and MiMo-V2-TTS models on March 19, 2026, targeting AI agents, multimodal processing and speech synthesis12.
$10 off the fastest Kimi K3 and GLM 5.2 from Synthetic
Model details
MiMo-V2-Omni
MiMo-V2-Omni is a Xiaomi omni-modal foundation model that combines strong multimodal perception with agentic capability in a single architecture. Rather than bolting on separate vision, video, and audio modules, it fuses dedicated encoders for each modality into one shared backbone so the model processes a unified perceptual stream rather than isolated inputs. This design lets the model see, hear, and read at the same time, an approach aimed squarely at real-world agents such as digital assistants executing multi-step software workflows, voice-controlled robotics, and autonomous systems that fuse sensor data on the fly.
The model was trained from the start to connect perception with action, learning to anticipate what will happen next and what should be done, so reasoning, tool use, and UI grounding emerge as a continuous process rather than a separate post-hoc stage. It was released on March 18, 2026 as part of a three-model MiMo-V2 lineup that also includes MiMo-V2-Pro and MiMo-V2-TTS, positioning the Omni variant as the multimodal entry point for agent-oriented workloads. In practice it is well suited to tasks that span modalities, from visual grounding and multi-step planning to function calling and code execution, making it a practical choice for builders who need one model that can both interpret rich real-world input and drive downstream actions.
Xiaomimimo-v2-omnimimodeprecated
Quick Info
Powered by- Provider
- Xiaomi
- Model key
- mimo-v2-omni
- Release date
- Mar 18, 2026
- Last updated
- Jun 24, 2026
- Knowledge cutoff
- 2024-12
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $0.14
- Output token cost
- $0.28
Limits
- Output tokens
- 131,072 tokens
- Context window
- 262,144 tokens
Latest news about MiMo-V2-Omni
Xiaomi
Coverage
The MiMo-V2 lineup includes three models: MiMo-V2-Pro, MiMo-V2-Omni, and MiMo-V2-TTS.
Xiaomi
Coverage
AIbase's coverage of Xiaomi's March 19, 2026 launch explicitly names MiMo-V2-Omni as the omni-modal base model that natively integrates text, vision, and audio, designed to connect perception to action execution within Xiaomi's agent stack. The report frames the launch as a coordinated trio release with MiMo-V2-Pro (1T Omni is positioned as the sensory layer that lets agents interpret real-world inputs and convert them into structured prompts for the Pro backbone, while pricing for the Omni variant is reported in line with the $1/M-token input tier used across the family. For technical readers, the AIbase piece is most useful for cla
Xiaomi
CoverageBenchmark
Benchmark scores and performance metrics for Xiaomi: MiMo-V2-Omni - MiMo-V2-Omni is a frontier omni-modal model that natively processes image, video, and audio inputs within a unified architecture. It combines strong multimodal perception with agentic capability - visual grounding, multi-step planning, tool use, and co
Xiaomi
CoverageBenchmark
Analysis of Xiaomi's MiMo-V2-Omni and comparison to other AI models across key metrics including quality, price, performance (tokens per second & time to first token), context window & more.