Gemini 3.7 flash vs deepseek v4 pro 0813 vs muse spark 1.2 – on voxel city dioramas
Three models each built three crossy road-style 3d scenes – a construction site, a nyc intersection, a river with a drawbridge – as single self-contained html files the setup: nousresearch's
I ran PureScript in Svelte (Deepseek did it but I'll take the credit anyways)
StateM: harness scaling for reliable agents
A runtime that gives agents durable state, checked transitions, and recoverable runbooks. On Terminal-Bench 2.1, it lifts GPT-5.5 to 92.1% and DeepSeek-V4-Flash to 88.1% for under $15.
Same prompt, 8 models. underwater voxel temple.
Qwen 27b local (8bit + 4bit), grok 4.6, swe 1.7, glm 5.2, luna, deepseek pro, deepseek flash. full prompt + side by side is here, don't take my word for the screenshots. look at
You have reached the end of the archive
All of deepseek