# DeepSeek 611 Hacker News commenters recommended DeepSeek and 244 criticized it, November 2025 to November 2026. It was recommended in 13 of 13 months; owners who said how long report a median of 0.2 years (19 owners). It meets the Hacker Approved badge rule. **Verdict:** People pick DeepSeek as a very cheap, open-weight model that comes close to frontier models for many tasks, but its censorship of China-sensitive topics and data-security concerns draw criticism. ## Why people recommend it - Extremely cheap to use (10 comments): "I just had claude teach me how to use it, and then tested it on the same example, which it got right, and for which it charged next to nothing." ([source](https://news.ycombinator.com/item?id=47954049)) "Chinese models cost a fraction of the price, and as someone who has been using DeepSeek and MiMo, they are nothing short of excellent." ([source](https://news.ycombinator.com/item?id=48747601)) "I've been using the deepseek api (not for coding though) and have been getting great results and it's so cheap it may as well be free." ([source](https://news.ycombinator.com/item?id=48122669)) - Near-frontier quality for most tasks (8 comments): "Any random startup can buy a fleet of GPUs for a few million and serve DeepSeek/GLM that are 85% as good as what OpenAI and Anthropic are spending hundreds of billions to accomplish." ([source](https://news.ycombinator.com/item?id=48682510)) "I switched to DeepSeek entirely once I decided to put 10 bucks on it and I realized that it could do whatever I was throwing at Claude or ChatGPT prior to that." ([source](https://news.ycombinator.com/item?id=49207506)) "several Chinese models (at least Qwen, DeepSeek, Kimi and GLM) are "close enough" that with the right harness they are cost effective alternatives" ([source](https://news.ycombinator.com/item?id=48390313)) - Open weights and published research (7 comments): "The best chinese models are deepseek (general purpose) and glm (coding) and they are both open weight and share lots of their tooling." ([source](https://news.ycombinator.com/item?id=48514532)) "open-weight LLMs like DeepSeek that prove the cost of running inference with near-frontier models can be very cheap" ([source](https://news.ycombinator.com/item?id=48106768)) "they are very open (not just weights). A young company and very much unlike Chinese Big Tech and American Big Tech." ([source](https://news.ycombinator.com/item?id=46123404)) - Strong coding with good harness (3 comments): "It took me around 15 minutes and a few dimes worth of DeepSeek tokens to upgrade some gnarly, janky, exploitable PHP 7.x code" ([source](https://news.ycombinator.com/item?id=50207654)) "if you combine the latest DeepSeek model with any good coding harness the results are surprisingly comparable to Claude Code" ([source](https://news.ycombinator.com/item?id=48150885)) "I love DS and am not looking back. For Elixir, not a lot of models are good, but DS just blazes through it." ([source](https://news.ycombinator.com/item?id=50132146)) ## What people warn about - Censors China-sensitive topics (8 comments): "a huge downside of Chinese-models is that these are chinese models with tianmen square and tibet and other issues" ([source](https://news.ycombinator.com/item?id=47128268)) "Not recognizing they were outputting wrongthink until after it was being streamed to the user is a known behavior with some Chinese chatbot apps." ([source](https://news.ycombinator.com/item?id=47187190)) "if you ask deepseek to deviate from the CCP stance that "Taiwan is an inalienable part of China that is in rebellion", it will become incoherent" ([source](https://news.ycombinator.com/item?id=46518577)) - Data security and CCP concerns (5 comments): "This is a risk although then this is fortunately a model that isn't tied to Chinese hosting. But indeed something to consider if using straight DeepSeek.com." ([source](https://news.ycombinator.com/item?id=48239791)) "giving your data to a non-deterministic program written by a geopolitical rival that has a long history of malicious intrusion on Western networks" ([source](https://news.ycombinator.com/item?id=50167235)) "Deepseek servers are CCP servers, there is no functional difference or any form of friction to keep the government "in check"" ([source](https://news.ycombinator.com/item?id=48242639)) - Weak at long agentic coding (3 comments): "Even Deepseek, which was and is revolutionary, was not useful for independent code commits longer than a few dozen lines." ([source](https://news.ycombinator.com/item?id=49570060)) "For chatbot-style usage maybe , can't really attest to that; for long-running, long-context "agentic" work including planning, within a harness etc., definitely not (yet)." ([source](https://news.ycombinator.com/item?id=50161962)) "every agent I've tried through Claude code (including deepseek and GLM) will actively try to avoid fixing pre-existing issues" ([source](https://news.ycombinator.com/item?id=48789208)) - Benchmarks overstate real-world ability (2 comments): "If true, would that sort of explain why Chinese Models score high on benchmarks, but not quite as capable when given real tasks?" ([source](https://news.ycombinator.com/item?id=49654439)) "A committee of good-not-great models (DeepSeek, MiMo, Gemma 4) cannot replicate the accuracy of Opus by itself." ([source](https://news.ycombinator.com/item?id=48483384)) ## Alternatives hackers recommend - [Claude Opus](https://hackerapproved.com/p/claude-opus.md): 3331 recommend - [OpenAI GPT](https://hackerapproved.com/p/gpt-5-5.md): 1205 recommend - [Claude Sonnet](https://hackerapproved.com/p/claude-sonnet.md): 854 recommend - [DeepSeek V4 Flash](https://hackerapproved.com/p/deepseek-v4-flash.md): 828 recommend - [Fable](https://hackerapproved.com/p/fable.md): 823 recommend ## Where to get it - [Visit the official site](https://www.deepseek.com/) ## Cite as Hacker Approved, "DeepSeek", https://hackerapproved.com/p/deepseek/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).