I am going to say something that many won’t like, but I honestly don’t see it playing out any other way
We are at a stage where no matter how much people push for “guardrails” and try to slow things down, the outcome is inevitable We’re on a path with AI that we can no longer
The best agent upgrade might not be a new model
It might be one of these 10 repos: 1) deepseek-harness Build the agent around plugins instead of hard-wiring everything Swap tools, interfaces or behavior without rebuilding the whole thing 2)
Upcoming deepseek v4.1 flash runs on ~300 tok/s
The preview endpoint still lacks vision, but I'm hoping the official release will be multimodal.
DeepSeek V4.1 Flash is a very interesting model.
I just tested it on the BridgeBench lava lamp test and it took LONGER to complete than Fable 5.1 and GPT 6 Astra. It ran at 344 toks/sec, spent 23.5M tokens with a cache hit rate of 99.7%, and cost $0.33. Even though it runs so
You have reached the end of the archive
All of deepseek