Meet Wally. Our inference stack for open frontier models, built to be the fastest place to run them.
Performance snapshot: GLM-5.3 Flash: 380 tok/s. GLM-5.3 Max: 790 tok/s Qwen3.8-27B: 485 tok/s DeepSeek-V4.1 Flash: 615 tok/s 1/7
I said let's try Deepseek Flash that fixes the well-known duck hunting game on the family computer "NES" Pictures from chatgpt 🦆
Quick commercial for deepseek
The first test for Grok 4.7 is completed and the result.....is not good!
I got a bit surprised with this one, as it is fairly "simple". I ask the model to create a complex diorama of "mini worlds". A city inside a coffee mug, an ocean inside a cup, a forest inside a book and a
You have reached the end of the archive
All of deepseek