DeepSeek V4.1 Flash + the Codex harness is a seriously underrated duo.
It COOKED this crash test. 🔥 I challenged it to simulate a small car crashing into TWO concrete mixer trucks… and a BATTLE TANK. 🚗💥 Watch the survival odds plummet. 💀 But here’s the real question: Does
Claude Chat GPT Gemini Deepseek
espiegleNov 810
Pelican Qin Shihuang Sun Wukong Polar Bear Test
Figure 1: Qoduck Figure 2: DeepSeek-V4.1-Flash-1017 Is the gap so big? ? ? DeepSeek is finished I’m very curious about what model Qoduck is. Put the prompt words in the comment area
The top of the coding leaderboards barely moved this month, but the middle got way more interesting.
On Arena WebDev, Kimi K3.1 Pro, Gemini 4 Argon, GPT-6.1 Terra and DeepSeek V4.1 Flash are all within about 30 points of each other, right around Opus 5 level. The difference is
You have reached the end of the archive
All of deepseek