My notes on Kimi-K2-Instruct-0905, aka Kimi K-2.1 – an incremental improvement on Moonshot's previous trillion parameter open weights model, now with twice the context length (256k up from 128k)
OPEN SOURCE
-
DeepMCPAgent: Dynamic MCP Tool Discovery and Agent Development
By
–
DeepMCPAgent A powerful tool for dynamic MCP tool discovery and agent development. Built with LangChain and LangGraph, it streamlines integration over HTTP/SSE while supporting major LLMs. Check it out! https://
github.com/cryxnet/deepmc
pagent
… -
OpenAI’s Journey from Open Research to Closed Commercial AI
By
–
> be openAI, circa 2017
> do lots of interesting research
> all open source > no one cares that much
> invent chatGPT in 2022
> get too busy to do open research
> everyone gets mad
> wheres the open research
> more like closedAI
> three years pass. now 2025
> finally release an -
Popular Coding Tool Created for Fun Powers AI Agents
By
–
It’s actually crazy how many coding agents use something I made for fun one afternoon
-
Efficient AI Model Runs on Less Than 200MB RAM
By
–
Good question! It's small enough to run on less than 200MB of RAM with quantization. You can read more about it here:
-
NanoBanana and EmbeddingGemma: Open AI tools for developers
By
–
Here's a quick look at some of the things we were up to this week —200M+ images have been edited with @NanoBanana share your favorites with us below! —We released EmbeddingGemma, an open model that helps developers build powerful on-device AI applications. Its efficient
-
Qwen3-Max-Preview Release Announced
By
–
BREAKING : Qwen got a big Qwen3-Max-Preview release. Now available via Qwen Chat & Alibaba Cloud API. Open models are conquering the space
-
RouteLLM API: Universal Gateway for All Language Models
By
–
RouteLLM API: one API for all your LLMs •Auto-routes to the best model for your prompt
•Or pick the model that you want
•Competitive open-source rates + closed-source failover This API is available with a ChatLLM subscription using your credits. -
FineVision: Free Open Dataset for Vision Language Models
By
–
We’re doing the work that nobody else wants to do! Welcome to FineVision, the best free open dataset to train vision language models. Let’s go open-source!
-
Moonshot updates Kimi K2 with 256k context
By
–
Moonshot updated Kimi K2 with a new “K2-0905” release that includes better coding performance and context length extension to 256k!