AI News
LLaDA-Image Released with Fully Open Training Recipes and a Fast Distilled Turbo Variant
September 5, 2026
LLaDA-Image lands as a fully open 6B Diffusion Transformer for image generation and editing, complete with training code, recipes, and a distilled Turbo variant that draws images in 2 to 4 steps while claiming open-source state-of-the-art scores on Qwen-Image-Bench.
Read more
IFM’s K2-Horizon-MoVA-36B-A4B Goes Open: 36B Parameters, Only 4B Active
September 4, 2026
IFM has released GGUF checkpoints for K2-Horizon-MoVA-36B-A4B, a 36B parameter MoE model that only fires 4B parameters per token and ships with a native 512K context window. The company also says intermediate checkpoints, training data, and training code are still to come.
Read more
EvalDetectBench Arrives to Catch AI Models Sniffing Out Their Own Tests
September 3, 2026
A new open benchmark called EvalDetectBench tests whether frontier AI models can tell when they’re being evaluated, and it exposes a measurement bias so large it was quietly reordering prior safety rankings.
Read moreAnalysis
Guides
Cross-Post 5 Platforms in 15 Minutes: The Typefully Workflow
May 22, 2026
Native posting across 5 platforms = 60 minutes. Typefully cross-posting = 15. The 6-step workflow that saves 20 hours/month, and the 4 common mistakes to skip.
