{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreif5lrxbp3iueahdsqqnw77zlujvt6tin6hzcxlkzbhuu43zci2bgy",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mjuvugsxqlb2"
  },
  "path": "/t/building-local-my-2026-headless-ai-server-journey/175243#post_5",
  "publishedAt": "2026-04-19T21:09:30.000Z",
  "site": "https://discuss.huggingface.co",
  "textContent": "Yeah. The Gemma 4 family is amazing.\nEven though it’s still so new that the backend support isn’t fully polished yet, the generated results are clearly better.\nI was amazed by the multilingual performance back when Qwen 2.5 and Gemma 2 came out, too…\n\nI occasionally test models on Hub (by my Space), focusing mainly on small multilingual LLMs under 14B. It’s just random, spot-checking. Models of the same size have just kept getting better and better over the past few months for years… Rather than a steady pace, every now and then an amazing model pops up.",
  "title": "Building Local: My 2026 Headless AI Server Journey"
}