Hunter Alpha first appeared anonymously on the AI platform OpenRouter on 11 March, sparking speculation that it was DeepSeek-V4, according to multiple reports. However, Xiaomi's AI model team MiMo confirmed that Hunter Alpha is an early internal test build of their MiMo-V2-Pro model.
The model identified itself as a Chinese AI model primarily trained in Chinese with a knowledge cutoff of May 2025, according to the Hunter Alpha chatbot. This cutoff matches that of DeepSeek's chatbot, multiple reports noted. When asked about its developer, Hunter Alpha declined to identify it, stating it only knows its name, parameter scale, and context window length, according to The Independent - Main. The model is described as a 1-trillion-parameter model with a context window of up to one million tokens.
I call this a quiet ambush - not because we planned it, but because the shift from chat to agent paradigm happened so fast, even we barely believed it.
Xiaomi's MiMo team is led by Luo Fuli, a former DeepSeek researcher, according to multiple reports. MiMo-V2-Pro is designed to serve as the 'brain' of AI agents and will partner with five major agent frameworks, including OpenClaw, offering a week of free access.
Following the announcement, Xiaomi's Hong Kong-listed shares surged by as much as 5.8% on Thursday, according to multiple reports.
People ask why we move so fast. I saw it firsthand building DeepSeek R1.
Separately, DeepSeek launched preview versions of its V4 model on Friday, including 'pro' and 'flash' versions with improvements in knowledge, reasoning, and agentic capabilities, according to DeepSeek. DeepSeek V4 is a successor to V3, released in late 2024. DeepSeek claims V4 Pro Max has superior performance relative to OpenAI's GPT-5.2 and Google's Gemini 3.0-Pro, but falls marginally short of GPT-5.4 and Gemini 3.1-Pro. DeepSeek also claims V4 pro outperforms Claude's Sonnet 4.5 and approaches Opus 4.5 in agentic capabilities. Both V4 pro and flash have a 1 million token context window, up from 128,000 in V3.
Some analysts remain skeptical about DeepSeek V4's performance, according to Ivan Su, senior equity analyst.
Unanswered questions remain about specific performance benchmarks for DeepSeek V4 compared to US rivals, how Xiaomi's MiMo-V2-Pro compares to DeepSeek V4, and the exact release date for DeepSeek V4.