Opus 5.2 holy shit get ready for this banger
everything in this image is code even the handwritings
in few hours i will drop the hardest one shot creation by any ai till date without burning crazy money
Opus 5.2 is coming and honestly it cannot come fast enough.
Opus 5 has been a complete disaster for Anthropic.
It was the best model in the world at launch. Then it got lazy, got slow, got nerfed, and lost to DeepSeek Flash on BridgeBench.
The model that was supposed to make Fable 5 unnecessary became the model nobody uses.
Anthropic needs Opus 5.2 more than we do at this point.
GLM 5.2 vs Qwen 3.7 Max.
Both mixture-of-experts. Both 1M context. Both built to run agents for hours.
The difference that actually matters: GLM 5.2 is open weight under MIT. Qwen 3.7 Max is API-only.
One you can own and self-host. The other you rent, for good.
Full breakdown, cost math, and how to run either on Yotta:
GLM 5.2 Fast is now available in Droid, hosted by @FireworksAI_HQ.
GLM 5.2 is one of the most popular models in Droid. Now you can build with the same quality at much higher speed.
GLM-5.2 just one-shotted a Minecraft-like game
Via Hermes, it generated a full voxel world with procedural terrain, water, basic building, and survival mechanics in about 20 minutes (~70k tokens).
Not perfect, but it’s a complete playable prototype made by an open model in one go.
Pretty impressive speed for this kind of thing.