Kimi K2 by Moonshot AI is a new 1T parameter open-source MoE model (32B active). It delivers state-of-the-art performance on coding, reasoning, and agentic tasks. Base and Instruct models are now available.
A 1T parameter open-source base model... with benchmark results this impressive. I really hope the delay of OpenAI's next open model isn't because they saw Kimi K2 coming🤔
But regardless, I know the Moonshot team has been putting in the work for this moment, and they absolutely deserve it. They've given the world an incredible open-source model. As far as I know, many dev communities are already incredibly impressed with its agentic coding capabilities.
A heads-up though: given this is a 1T parameter model, running it locally on consumer hardware will be tough for most. The easiest way to get started is by Kimi's official website or API.
Report
Running Kimi K2 locally needs very powerful hardware, like top-tier GPUs or expensive computers, which most people don’t have. How do you plan to make Kimi K2 more accessible for users with regular hardware?
@vouchy I don't think this is the point. It's more that you have access to it and have full confidence that this is the model you're using, instead of self-hosting a 1T parameter (32b active) LLM on your own.
That's why we have services like OpenRouter, and even the official API from Moonshot.
Report
Congratulations on the release Zan and team. Good luck!
Incredible work, Moonshot team! A 1T parameter MoE model stepping into the arena feels like a true dharma-yuddha of innovation. Kimi K2’s leap in coding and reasoning benchmarks shows not just scale, but thoughtful engineering. May this open-source offering inspire both progress and ethical wisdom as we chart the next frontier of agentic AI. Excited to see how the community builds upon this luminous foundation! 🚀
Report
Kimi K2's impressive performance in coding and agentic tasks is a great breakthrough for the open - source AI community! For developers who want to fine - tune Kimi K2 for specific applications, are there any detailed documentation or tutorials available to guide them through the process?
K2 has been the most popular topic thses days, really a strong model especially working with Claude Code. A cheaper and more powerful combo and choice.
Report
We love new independent models. Good luck with the launch today!
Replies
Flowtica Scribe
Hi everyone!
A 1T parameter open-source base model... with benchmark results this impressive. I really hope the delay of OpenAI's next open model isn't because they saw Kimi K2 coming🤔
But regardless, I know the Moonshot team has been putting in the work for this moment, and they absolutely deserve it. They've given the world an incredible open-source model. As far as I know, many dev communities are already incredibly impressed with its agentic coding capabilities.
A heads-up though: given this is a 1T parameter model, running it locally on consumer hardware will be tough for most. The easiest way to get started is by Kimi's official website or API.
Running Kimi K2 locally needs very powerful hardware, like top-tier GPUs or expensive computers, which most people don’t have. How do you plan to make Kimi K2 more accessible for users with regular hardware?
SpotVault
@vouchy I don't think this is the point. It's more that you have access to it and have full confidence that this is the model you're using, instead of self-hosting a 1T parameter (32b active) LLM on your own.
That's why we have services like OpenRouter, and even the official API from Moonshot.
Congratulations on the release Zan and team. Good luck!
Fish Audio
INSANE
Incredible work, Moonshot team! A 1T parameter MoE model stepping into the arena feels like a true dharma-yuddha of innovation. Kimi K2’s leap in coding and reasoning benchmarks shows not just scale, but thoughtful engineering. May this open-source offering inspire both progress and ethical wisdom as we chart the next frontier of agentic AI. Excited to see how the community builds upon this luminous foundation! 🚀
Kimi K2's impressive performance in coding and agentic tasks is a great breakthrough for the open - source AI community! For developers who want to fine - tune Kimi K2 for specific applications, are there any detailed documentation or tutorials available to guide them through the process?
AltPage.ai
K2 has been the most popular topic thses days, really a strong model especially working with Claude Code. A cheaper and more powerful combo and choice.
We love new independent models. Good luck with the launch today!
SpotVault
It's amazing what foundations y'all used, already added it to shiori.ai