DeepSeek just gave their fast agent model eyes, so it can look at a screenshot and build from it.
Before this, it was text only. You couldn't show it anything. Now the image goes straight into the reasoning, not through another model that describes it first. That matters.
I tested Gemini 3.7 vs DeepSeek V4 Flash on a 5-phase CTF challenge!
🎯 🏆 Gemini 3.7: Solved all 5 phases and retrieved the flag. ❌ DeepSeek V4 Flash: Got stuck in the multi-step logic. 🚀 Gemini 3.7 wins this round!
[Shocking] For those who want to reduce Claude's charges. The name is clawcodex, and the execution cost is 1/230th of the original.
Full scratch port with over 230k lines, released under MIT license For those who have been paying monthly, this is a normal incident 😳 ・Clawcodex, completely free at MIT ・Execution cost is 1/230 of the original ・DeepSeek cache works as is
When looking at X announcements, each company flaunts its own Flash model
When looking at X announcements, each company flaunts its own Flash model: "surpasses Opus 4.8, close to Fable 5", but no one knows what the actual results will be! So, I organized four mainstream Flash models and held an F1 ranking competition. I refused to allow manufacturers to unify the gods.
You have reached the end of the archive
All of deepseek