Top-5 Best Value #LLM Models of the Day at UTC-04
| Model | #AAII | Price | | GLM 5.3 Flash | 42 | $0.09 | | DeepSeek V4.1 Flash | 40 | $0.19 | | Gemini 3.8 Flash | 41 | $0.57 | | DeepSeek V4 Flash 0731 | 34 | $0.06 | | GPT-5.6 Luna | 38 | $0.18 |
I just read the Anthropic report and was not scared by “AI.”
This scared me: an army is no longer needed to do state work. In 8 months they documented Claude in cyberattacks, surveillance, scams, propaganda, illegal copying of models and software linked to weapons. Russia
AI overtakes in curves and rolls over very seriously. Anthropic is a powerful fisherman
AI overtakes in curves and rolls over very seriously. Anthropic is a powerful fisherman, accusing China's big models of not only distilling Claude, but the most terrifying thing is that when Kimi and DeepSeek are used by state-owned enterprises and DeepSeek, they forward users' conversations to Claude, and then output Claude's answers to their own model users.
The first is a Pelican bike I built using Grok 4.6 xhigh from DeepSeek Harness
The first is a Pelican bike I built using Grok 4.6 xhigh from DeepSeek Harness, the second is a Grok 4.6 xhigh from Cursor, and the third is a Grok 4.6 xhigh from Grok Build. Overall conclusion, don’t use it for now, grok will sit at the same table as gemini before 4.7. No, sit at the same table with Wen Xinyiyan. elonmusk…
You have reached the end of the archive
All of deepseek