{"slug": "selfhosted-ai-server", "title": "Selfhosted AI Server", "summary": "A forum user asked for recommended specs for self-hosted AI workloads with Odysseus, and a respondent said a 3090 Ti should suffice, noting that a Qwen3.8-27b quant fits in its VRAM and is a good small model, with context window size determining whether an AWQ or flat 4-bit quant fits.", "body_md": "[AnHero](https://forum.level1techs.com/u/AnHero)\n1\n \nHello, I’m new to self hosted AI workloads and wanted to know what the recommended specs for decent performance with something like odysseus. I currently have a 30090ti to spare but not sure if that’s gonna be enough. Thanks in advance!\n\n \n\n \n3090 Ti should do fine. A Qwen3.8-27b quant should fit in the VRAM on that and its a wickedly good small model. Whether you can fit an AWQ quant or a flat 4-bit depends on how large a context window you need. I haven’t fooled with Odysseus so I don’t know what it specifically needs.", "url": "https://wpnews.pro/news/selfhosted-ai-server", "canonical_source": "https://forum.level1techs.com/t/selfhosted-ai-server/255536#post_2", "published_at": "2026-09-09 19:18:32+00:00", "updated_at": "2026-09-09 20:13:25.920945+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-infrastructure"], "entities": ["Odysseus", "Qwen3.8-27b", "3090 Ti"], "alternates": {"html": "https://wpnews.pro/news/selfhosted-ai-server", "markdown": "https://wpnews.pro/news/selfhosted-ai-server.md", "text": "https://wpnews.pro/news/selfhosted-ai-server.txt", "jsonld": "https://wpnews.pro/news/selfhosted-ai-server.jsonld"}}