Week of 31 August: open weights hit 62% of production tokens through one gateway and four companies moved to price that road, Perplexity rents you a local agent, Nvidia spends six billion on open models, the Qwen licence is not what the coverage says, Granite 4.2 puts the thinking dial in your hands, and a song gets printed on paper.
Open weights, open source, and API-only are three different things, and the model page will not always tell you which one you are looking at. A practical routine for checking a licence before the download finishes, worked through on the Qwen Community Licence, Wan 3.0, and Apache 2.0.
How to animate an image using AI on your own GPU, the one-beat rule, real-time-speed camouflage, anchor plates, and when to rent a cloud GPU instead. The Engine Room, machine three.
Our 66-second monster movie, helicopter, dragons, a meta twist, cost $16 including every failed take. The full filmmaking pipeline: storyboards, cloud + local animation, score, assembly.
We produced a 29-second helicopter-and-dragons action tease with image-to-video AI. Here’s every mistake we made, tangled limbs, teleporting wardrobes, shrinking monsters, and the directing rules that fixed them.
A working comparison from our production pipeline: 24-second local renders on an RTX 5090 vs 15-second cloud generations, cost, speed, quality, and when we actually use each.