agreed
agreed
More posts
A web-based guitar created using DeepSeek-V4.1-Flash
; no Three.js was used.
DeepSeek V4.1 Flash beta is open.
Model ID: deepseek-v4.1-flash-expires-on-0910 Same base_url, just swap the model name. New architecture — native multimodal, images and text in one model now. No more Vision Encoder bolted on. Same price as V4 Flash: $0.14/1M input, $0.28/1M output.
Meituan shipped LongCat-2.5
. 1.6T params, ~48B active, native multimodal, aimed at long-horizon agents, trained start to finish on Chinese hardware. Images get quantized into multi-level discrete tokens, and that's where Visual Coding and 3D pick up. On the RL side, failed trajectories get turned back into verifiable training tasks. W4A4 inference, 43% faster prefill. Meituan's own evals put it between Claude Opus 4.8 and Opus 5.
SenseTime open-sourced SenseNova Loop-9B-Preview.
It stacks layers instead of parameters. The decoder has 32 layers, but layers 4 through 27 run twice with shared weights, so a forward pass does 56 layers of work while the checkpoint stays at 9B. Their own eval numbers: 71.38 on BIRD-Dev and 13.72 on AutomationBench, both above Qwen3.5-27B. Terminal Bench is still behind.
