{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreie2szq6vrh73l5cladbtulh774ektzk22b3acg7soqo6se52tprbq",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mh5wccg42nd2"
  },
  "path": "/t/are-there-any-llms-that-can-run-with-decent-performance-on-hardware-comparable-to-jetson-nx/174305#post_1",
  "publishedAt": "2026-03-16T06:09:49.000Z",
  "site": "https://discuss.huggingface.co",
  "textContent": "Hi, I’m working to make an AI agent for robotics application.\n\nWe plan to use Jetson AGX or NX, and want to run all the models locally if available, without using cloud or on-premise server. But then I found out that the memory bottleneck is too harsh. I have tried several models including Llama, Qwen, Gemma (7~9B), but they were too slow. Some of them even did not run.\n\nIs there any language model that can solve my problem? The agent would be used for various tasks including on-site diagnosis, production of operator-ready summaries, and more.",
  "title": "Are there any LLMs that can run with decent performance on hardware comparable to Jetson NX?"
}