Sources told CNBC that Moonshot AI tried to lease access to Nvidia Blackwell clusters in Europe after U.S. officials raised concerns about remote access.
Why it matters: Remote access to overseas GPU clusters can sidestep chip export limits, creating compliance exposure for firms that rent offshore compute.
2 sources
Artificial Analysis launched VIALS-AA, a benchmark for scientific image reasoning, where Gemini 4 Carbon (high) leads with 64% of tasks answered correctly in all three repeats.
Why it matters: A single-pass score can hide large differences in reliability and cost on image-based scientific work, where a result that changes between runs can send teams toward wrong conclusions.
1 source
Moonshot released weights for its K3.1 Pro flagship, a 2.8T-parameter model with 104B active parameters and a 1M-token context.
Why it matters: A model of this size with open weights gives companies a self-hostable option at frontier scale, though the hardware needed to run it is substantial.
2 sources
A 50-task benchmark of games and web tasks shows top models pass only 36% of tasks under real-time latency, versus up to about 80% with a paused clock.
Why it matters: The gap between paused and real-time results suggests current agents are unreliable wherever response speed matters, such as buying tickets or bidding in auctions. Teams should not assume that strong offline benchmark scores carry over to live, latency-sensitive workflows.
1 source
Reports say DeepSeek and Moonshot AI use GPT-6 Astra outputs to train new models after Anthropic tightened defenses against distillation.
Why it matters: Model-output theft is emerging as a competitive and security issue, which could lead to stricter access controls and terms of service across frontier model providers.
1 source