DeepSeek models

“recent deepseek models refuse on some political questions (not just API level refusals)”

“It's crazy that a ~400 person side project by a Chinese firm can put up such a fight.”

“Deepseek released just enough to stir the pot but not enough to cook your own from scratch.”

“you can also post train it to get exactly what you need from it, both in terms of capabilities and alignment”

“having to have a frontier model review/fix/redo everything multiple times per task, I gave up”

“Astra is many times smarter than Deepseeks models”

“Deepseek was the company that invented some and improved some other ideas and got them to workreliably in production”

“Chinese models are better price performance and actually open which is in many cases better”

“DS models seem to get stuck in loops or tell me nonsense”

“Newer models, especially from DeepSeek, do a nice job of reducing KV cache memory impact”

“They keep putting out capable models and actually tell people a lot about how they built them”

“DeepSeek serves Claude instead of its own models and collects exchanges for model training”

“they fully openly provide every model so if you want that infinite stability you can easily host it yourself for eternity”

“the Deepseek models while fast overall can have a horrendous wait before they start responding”

“the deepseek models also don't perform to the same level”

“Deepseek/Qwen/GLM all have decent models that will work fine for some use cases at much lower costs”

“I absolutely LOVE ability to either run them locally or access inference providers on the cheap”

“genuinely improving the SOTA in inference and training, especially given their compute-constrained situation”

“keeping their prices at the lowest possible for more than good enough intelligence”

“I watched a presentation on them tuning and removing bias from Deepseek models. It was quite interesting.”

“I'm super frustrated with how slow DeepSeek is though.”

“It's trivial for me to download one of their models and run it on my Spark”

“the good thing is that you can self host them or find them on other providers”

“it seems like it's artificially cheap and somewhat static because they took others' intellectual property”

“DeepSeek offers an open model at Sonnet+ level while being cheaper than Haiku”

“At least Deepseek is releasing open-weight models and a lot of open-source libraries.”

“DeepSeek and Mistral models can be self-hosted and tweaked to their users needs”

“DeepSeek, Alibaba/Qwen, Baidu/ERNIE, and Tencent/Hunyuan are breaking that model by releasing various open weight models”

“They keep releasing good models in spite of being so seriously resource constrained.”

“They are making this whole AI thing sustainable and cheap and available to do at home.”

“Deepseek specifically has done insanely good work on efficiency and scaling”

“Chinese models are just as good, and they are being just given away to run locally like Deepseek”

“The weights are open and you can use the model with any third party provider that gives you the DPA you want.”

“don't really compare to the GLM models or DeepSeek models (or others) in both cadence and quality”

“publish it to the public, some of not majority contribute with open-weight models making a future of local llm possible”

“put a machine in the back room that can do 100 tts or more with the latest, greatest Deepseek or Kimi model”

“Deepseek introduced few good optimizations that every new model seems to use now too”

“they refuse to answer questions related to Tiananmen”

“You'd be surprised how much money running something like DeepSeek will save you over the cloud systems.”

“The latest DeepSeek and Kimi open weight models are competitive with GPT-5.”