• Top
  • New

DeepSeek v3 beats Claude sonnet 3.5 and way cheaper

by helloericsf on 12/26/2024, 11:47:29 AM with 9 comments
  • by helloericsf on 12/26/2024, 11:48:43 AM

    HF link: https://huggingface.co/deepseek-ai/DeepSeek-V3 Aider link: https://aider.chat/docs/leaderboards/ Pricing($0.14/$0.28 per 1M tokens) reference:https://x.com/xingyaow_/status/1872145835699691675?ref_src=t... LiveBench via reddit: https://www.reddit.com/media?url=https%3A%2F%2Fpreview.redd....

  • by patrickhogan1 on 12/26/2024, 2:17:45 PM

    It does not beat Claude Sonnet 3.5 on SWE Bench (42 to Claude's 50). It chooses 4 benchmarks of the 100s of available benchmarks and then decides it "beats" Claude Sonnet 3.5.

  • by Jet_Xu on 12/30/2024, 7:40:11 AM

    Please refer to my recent AI Code review performance test include DeepSeek V3: https://news.ycombinator.com/item?id=42547196

  • by sam_goody on 12/26/2024, 5:17:49 PM

    What are the minimum and recommended amounts of RAM, hard disk space, CPU or GPU to run this locally.

    As someone who just follows this stuff from afar, it is hard for me to conceptualize if this is a SaaS only model, or if it means we are getting to the point where you can have a A1 model on a local machine.