Log inSign up
Image
user avatar
am.will
@LLMJunky
StarSwap // Lynk // DevX Director of n number of agents Also not a car. amwill.dev
Texas
star-swap.com
Joined December 2025
2,336
Following
28.4K
Followers
RepliesArticlesMedia
  • Pinned
    user avatar
    am.will
    @LLMJunky
    May 29
    Introducing Lynk, a brand new way to interact with your favorite harnesses on the go. Compatible with OpenClaw, Hermes, Codex, and local edge models. Fully featured client allowing you to easily and quickly switch between your favorite agents. Lynk has been my absolute
    Image
    00:00
  • user avatar
    am.will
    @LLMJunky
    1h
    I've seen quite a few people say they were disappointed by the Qwen 3.8 Max release, and I get it...but that's literally only because Kimi K3 exists. Qwen 3.8 27B doesn't have a Kimi K3 to worry about. The math is simple: their 3.6 version is still arguably the best small
    Image
  • user avatar
    am.will
    @LLMJunky
    1h
    anything under deepseek v4 flash basically doesn't exist to me. with <200GB of vram, i'm getting ~800K context @ 200 tokens per second. It's not quite Cerebras fast, but damn. It is a joy to use, and its basically somewhere in the opus 4.7/4.8 range. It's incredibly good at
    Image
    00:00
  • user avatar
    am.will
    @LLMJunky
    3h
    In exactly a day and a half, Qwen 3.8 27B will absolutely DOMINATE consumer devices. Prediction: it will have ~Opus 4.6 level performance on a single 3090, exactly as @TheAhmadOsman predicted last year. Think about that. A local model running on a single GPU able to do things
    Image
  • user avatar
    am.will
    @LLMJunky
    6h
    <rant> I want to see what model & reasoning level my subagents are using! Claude app should nvrrr obfuscate this from users </rant>
    Image