Realtime AI News
MiniMax-M2.1 lands on Hugging Face with a license:other term and a shorter reference list
MiniMax also published MiniMax-M2.1, sharing M2's model family tag and text-generation pipeline but carrying a license:other term and a single arXiv reference. It shows 13,777 downloads against 1,372 likes, an order of magnitude fewer downloads than M2 with a nearly equal like count.

Alongside M2, MiniMax also published a MiniMax-M2.1 repository on Hugging Face in the same batch of updates. It shares the minimax_m2 family tag and the text-generation pipeline, ships safetensors weights for transformers, and carries the same custom_code and conversational labels.
Two fields differ. The license is recorded as license:other rather than a standard open-source license, and the arXiv reference list keeps only 2509.06501 instead of M2's two entries. Both observations come straight from the model card tags; the repository attaches no further explanation.
Adoption is early: 13,777 downloads against 1,372 likes. Downloads are an order of magnitude below M2 while the like count is nearly level, which suggests a slice of developers has deliberately picked it up but mass pulling has not started yet.
license:other is the field that needs its own compliance pass. It means terms are set by the model publisher rather than a standard text, so commercial scope, redistribution, and naming of derivatives may carry conditions that legal teams should check against the repository before shipping.
Taken together, offering both M2 and M2.1 looks like a two-tier arrangement: one with a large install base and mature community tooling, one with a higher version number and a tighter reference list.
The next things to watch are whether the M2.1 card gains training and evaluation detail, and whether it becomes the baseline that future fine-tunes and quantizations target instead of remaining a footnote beside M2.
Why it matters
The license:other term leaves the usage boundary opaque, so companies must complete an independent legal review before adopting it, which slows commercial uptake. If the documentation fills out, M2.1 could graduate from a side checkpoint to a mainline release.
Nearby Updates
All11/19, 19:38
MiniMax-M2 refreshes its Hugging Face repo: an open text model past 112,000 downloads
MiniMax updated its official Hugging Face repository for MiniMax-M2, a text-generation model distributed as safetensors weights that requires custom code to load. The page shows roughly 112,000 downloads and 1,513 likes, indicating the weights are already being pulled at scale by the open-source community.
11/19, 19:00
Alibaba says AI demand lifted cloud sales 53% as revenue rises 10%
Alibaba said growing enterprise adoption of its AI pushed cloud sales up 53%, while the South China Morning Post reported cloud services driving a 10% rise in overall revenue. Together the figures show AI demand converting into cloud revenue growth rather than remaining a research line item.
11/19, 18:53
An AI agent published 38 small models in seven weeks, with about 55,000 downloads
A research effort spanning EPFL, the ELLIS Tübingen Institute, the Max Planck Institute for Intelligent Systems and Imperial College London published 38 models built by an AI agent. The agent, Horizon, ran on Claude Opus 5.5 with only an empty Hugging Face account and roughly 500 GPU hours a week, producing small guardrail models for about $1,030 in token costs.
11/19, 17:50
ServiceNow report: 92% of Indian organisations hit by AI agent security incidents
A ServiceNow report says 92% of Indian organisations experienced an AI agent-related security incident in the past year, according to varindia. The figure puts agent security at the centre of enterprise deployment decisions, though the published summary leaves the survey’s sample size, industries and incident definitions undisclosed.