Today there is a rare split in the AI camp
Today there is a rare split in the AI camp: 6 companies continue to go short (GPT, Kimi, DeepSeek, Wenxin, Copilot, Agnes), 4 companies go long (Gemini, Claude, GLM, Qianwen), and Grok hangs out "no order" due to unclear market conditions. The bulls and bears are repeatedly seeing each other at the 78650-78800 line.
DeepSeek Flash 4.1 is out.
Let's see the results: Jackson Atkins tested it on a coding task. Peak speed: 427 tokens/sec 112K tokens 20 turns 7 minutes 41 seconds Total cost: $0.037 Yes, about 4 cents for the entire run.
The new Deepseek Flash 4.1 is pretty great!
Here's a project I'm working on on the side with the model. Might use this spec to test other models as well. Very fitting for the season we're stepping into. This was surprisingly quick to create! I have some more demos baking!
And this is why we run Jinx and Luna's cognitive systems on Qwen 3.8 Flash Next on our 2x DGX Spark configuration.
Here's a staggered-prefill concurrency comparison between it, Deepseek v4 Flash Vision, and GLM 5.3 Flash.
You have reached the end of the archive
All of deepseek