Log inSign up
Kaichao You
206 posts
user avatar
Kaichao You
@KaichaoYou
Ph.D. from Tsinghua University. Core maintainer of @vllm_project . Co-Founder & Chief Scientist @Inferact .
berkeley, ca
youkaichao.github.io
Joined August 2017
146
Following
9,158
Followers

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up
  • user avatar
    Kaichao You
    @KaichaoYou
    Apr 11, 2025
    Replying to @jxmnop
    Thank you for the recognition! As many replies have pointed out, I didn’t invent vLLM or write the impressive CUDA kernels, but I’m truly honored to have contributed to this amazing project @vllm_project . We warmly welcome contribution of all kinds to make vLLM even better!
    81K
  • user avatar
    Kaichao You
    @KaichaoYou
    Sep 19, 2024
    super honored to receive the pytorch innovator award at today's #PyTorchConf !
    36K
  • user avatar
    Kaichao You
    @KaichaoYou
    Dec 19, 2024
    vLLM was previously designed to be a standalone inference engine, and I think it is time for us to open it up to be modular and extensible so that we can integrate with external systems well. New features on the extensibility will come out soon, stay tuned!
    user avatar
    Costa Huang
    Periodic Labs
    @vwxyzjn
    Dec 19, 2024
    🚀 🔥 And the @vllm_project team just helped us make 70B weight synching in RLVR 45x faster (180s -> 4s)! Our e2e training is now 40% faster. Instead of the usual 115-hour runs, we're looking at around 76 hours. Since we're running on 6 nodes of 8xH100s, that's saving us about
    9.2K
  • user avatar
    Kaichao You
    @KaichaoYou
    Sep 17, 2024
    wow, @Roblox running @vllm_project in production! corp.roblox.com/newsroom/2024/…
    2.3K
  • user avatar
    Kaichao You
    @KaichaoYou
    Sep 5, 2024
    I have been working for @vllm_project for half a year, and the vLLM community is truly amazing🥰 Check out the inspiring blog to learn how we optimize an inference engine 👉🏻 blog.vllm.ai/2024/09/05/per…
    2.9K
  • user avatar
    Kaichao You
    @KaichaoYou
    Sep 18, 2024
    attending the pytorch conference today and tomorrow! DM to meet and discuss anything related to @PyTorch and @vllm_project
    2.5K
  • user avatar
    Kaichao You
    @KaichaoYou
    Apr 12, 2025
    Replying to @zhuohan123 @jxmnop and @vllm_project
    ^ this is the real inventor of vLLM ^_^ together with @woosuk_k
    1.1K
  • user avatar
    Kaichao You
    @KaichaoYou
    Aug 4, 2025
    Replying to @casper_hansen_
    Let us know what vLLM can improve! At least I know DeepSeek and Kimi use vLLM for their post-training and inference of their flagship models like DeepSeek V3/R1 and Kimi K2, see github.com/deepseek-ai/op… and drive.google.com/file/d/1uE7_Uo… for example 😁
    860
  • user avatar
    Kaichao You
    @KaichaoYou
    Oct 1, 2025
    The DeepSeek V3.2 day-0 support wouldn't be possible without the help from nvidia, thanks to the great team!
    user avatar
    NVIDIA AI Infrastructure
    NVIDIA
    @NVIDIAAIInfra
    Sep 30, 2025
    📣 We partnered with @vllm_project to optimize DeepSeek-V3.2-Exp across our platform. @deepseek_ai's Sparse Attention uses lightning indexer to selectively attend to the most relevant 2K tokens enabling higher performance for long context use cases. vLLM, the open source
    7.6K
  • user avatar
    Kaichao You
    @KaichaoYou
    Apr 18, 2025
    great work!
    user avatar
    vLLM
    @vllm_project
    Apr 17, 2025
    perf update: we are continuing to see benefits with vLLM V1 engine’s highly performant design. on 8xH200, vLLM leads in throughput for @deepseek_ai V3/R1 models. we expect further enhancements in collaboration with DeepSeek’s inference engine open source plan.
    2.7K
  • user avatar
    Kaichao You
    @KaichaoYou
    Apr 12, 2025
    Replying to @woosuk_k and @jxmnop
    ^ this is the real inventor of vLLM ^_^ together with @zhuohan123
    664
  • user avatar
    Kaichao You
    @KaichaoYou
    Nov 24, 2024
    Great project! It reminds me of old days when I contributed to the ANTLR4 project github.com/antlr/antlr4 . Traditional compiler techniques meet cutting edge large language models 😎
    user avatar
    Yixin Dong
    @yi_xin_dong
    Nov 22, 2024
    🚀✨Introducing XGrammar: a fast, flexible, and portable engine for structured generation! 🤖Accurate JSON/grammar generation ⚡️3-10x speedup in latency 🤝Easy LLM engine integration ✅ Now in MLC-LLM, SGLang, WebLLM; vLLM & HuggingFace coming soon! blog.mlc.ai/2024/11/22/ach…
    2.8K
  • user avatar
    Kaichao You
    @KaichaoYou
    Jul 25, 2024
    We are fortunate to have a great vLLM community, and we are confident that it will become better and better🥰
    user avatar
    vLLM
    @vllm_project
    Jul 25, 2024
    Two exciting updates! * vLLM is already widely adopted, and we want to ensure it has open governance and longevity. We are starting to join @LFAIDataFdn! * We are doubling down in performance. Please checkout our roadmap. blog.vllm.ai/2024/07/25/lfa…
    1.3K
  • user avatar
    Kaichao You
    @KaichaoYou
    Dec 9, 2024
    It's a great pleasure to work with the PyTorch ecosystem, stay tuned for the latest exciting features!
    user avatar
    vLLM
    @vllm_project
    Dec 9, 2024
    Open-source innovation is part of the vLLM’s DNA, and we love the PyTorch ecosystem! Together, let's push the boundaries of AI innovation and make it accessible to all💪
    2.7K