Astra, DeepSeek V4.1 cloud and local compared on the same water prompt
On one water-simulation prompt, Astra finished in 30 seconds, DeepSeek V4.1 cloud took 103 seconds and 28.5K tokens, and the local setup on four DGX Sparks took 25 minutes.
Astra-6 vs DeepSeek-v4.1 cloud vs local. Same water prompt. Left → right: • Astra: “Worked for 30s” in ChatGPT • DeepSeek V4.1 Cloud: 103s, 28.5K completion tokens • DeepSeek V4.1 Local: 4× DGX Spark, 25m to HTML through Hermes 512×512 mesh. Click-to-ripple. Interacting waves. One HTML file. Might be just me but deepseek did a way better/more interesting render Astra: https://stillwater-wave-lab.rwesche.chatgpt.site/ Cloud: https://wesche.com/lab/deepseek4.1cloud/waterthreejs/ Local: https://wesche.com/lab/deepseek4.1local/waterthreejs/