Weeks ago I wrote a post saying the next big AI leap would not come from San Francisco. It would come from a mobile company. MiMo V2.6 just dropped. And it proved exactly what I was talking about.
They are pushing reinforcement learning harder than anyone right now. The new model is literally focused on RL. They even livestreamed the training run. People are already testing it and saying it is a top tier alternative for agentic coding.
They kept the API pricing exactly the same as V2.5. They did not raise the price. Why would they. They do not just want your money. They want your data. They want to keep the firehose of pure reinforcement learning gold flowing into their servers.
The big labs are burning cash. Xiaomi is playing a completely different game. You can read my original breakdown where I predicted this right here:
Top comments (0)