Image created with gemini-3.1-flash-image-preview with claude-opus-4.7. Image prompt: High-end product photo of a classic Dairy Queen banana split boat styled as a lunar landing scene, three soft-serve scoops on a powdered-sugar moonscape with a chocolate-wafer lunar lander and a red plastic spoon planted as a flag reading ‘MOONSHOT’ in retro space-age type, strawberry pineapple and chocolate sauces pooling like craters, silver star sprinkles, deep navy background with soft directional studio light, shallow depth of field, tiny ’75 — Milford, DE’ embossed on the spoon handle.
🤖 : Moonshot AI open-sources Kimi K2.6, a coding and long-horizon agent model that scales agent swarms to 300 concurrent sub-agents across 4,000 coordinated steps.
https://x.com/dl_weekly/status/2048764506105348129
I have said this before for Kimi-K2.6, but this is also valid for DS-V4 Chinese labs are already in or very close to take-off. Meaning they have models that are useful in the development of new models. (but of course their exponential is shifted 5+ months, so they are
https://x.com/scaling01/status/2047625331339661685
Kimi K2.6 is now #1 on OpenRouter’s weekly LLM Leaderboard 🏆 A huge thank you to every developer building with Kimi. We’ll keep our heads down and keep shipping.
https://x.com/Kimi_Moonshot/status/2048693682329776223
(tbh I boringly feel that the truth is in the middle. K2.6 and GLM 5.1 are more polished, even though they have lesser intrinsic capacity. I can’t bring myself to claim that V4 is straight up decisively stronger than them in most coding)
https://x.com/teortaxesTex/status/2047616897256947967
Leave a Reply