{"slug": "ninfer", "name": "ninfer", "category": "ai", "category_name": "AI models and assistants", "description": "", "recommenders": 3, "commenters": 3, "warners": 0, "score": 2.874, "rising": 3.0, "first": "2026-08-14", "latest": "2026-09-03", "why": "Strongly recommend https://github.com/Neroued/ninfer, which can pull ~180 TPS on 5090 with 3.8", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 3, "neutral": 0, "compare": 0, "warn": 0}, "spark": [0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 1, 0], "listed": true, "loved": [], "caveat": "", "short_quote": "Strongly recommend https://github.com/Neroued/ninfer, which can pull ~180 TPS on 5090 with 3.8", "icon": true, "domain": null, "github_repo": "github.com/neroued/ninfer", "official_url": "https://github.com/Neroued/ninfer", "pricing": "", "aliases": ["ninfer"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 792884672, "item_id": 49555292, "url": "https://news.ycombinator.com/item?id=49555292", "by": "pllbnk", "date": "2026-09-03", "stance": "recommend", "quote": "on RTX 5090 I can now get ~200 tok/s and over 400 tok/s on concurrent requests which is plenty fast for a local model", "excerpt": "Just a couple days ago I learned about ninfer (https://github.com/Neroued/ninfer) and on RTX 5090 I can now get ~200 tok/s and over 400 tok/s on concurrent requests which is plenty fast for a local model of this strength.", "context": "", "use_case": "local LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 790958528, "item_id": 49434908, "url": "https://news.ycombinator.com/item?id=49434908", "by": "Philpax", "date": "2026-08-25", "stance": "recommend", "quote": "Strongly recommend https://github.com/Neroued/ninfer, which can pull ~180 TPS on 5090 with 3.8", "excerpt": "Strongly recommend https://github.com/Neroued/ninfer, which can pull ~180 TPS on 5090 with 3.8, and 500 (!) with 3.6 35B-A3B.", "context": "runs on an RTX 5090", "use_case": "fast local LLM inference", "switched_from": "", "story": "Qwen 3.8-Flash-Next releasing tomorrow (125B a6B)"}, {"id": 788895008, "item_id": 49305938, "url": "https://news.ycombinator.com/item?id=49305938", "by": "kimsey0", "date": "2026-08-14", "stance": "recommend", "quote": "as inference engine gets me ~138 tokens/second, roughly double what I get with a naive llama.cpp setup.", "excerpt": "If anyone else is running this on an RTX 5090, https://github.com/Neroued/ninfer as inference engine gets me ~138 tokens/second, roughly double what I get with a naive llama.cpp setup.", "context": "runs an RTX 5090", "use_case": "local inference on RTX 5090", "switched_from": "llama.cpp", "story": "Qwen 3.8 27B is out: open weights, best local dense model yet"}]}], "caveats": [], "who": [{"by": "Philpax", "context": "runs on an RTX 5090", "url": "https://news.ycombinator.com/item?id=49434908", "date": "2026-08-25"}, {"by": "kimsey0", "context": "runs an RTX 5090", "url": "https://news.ycombinator.com/item?id=49305938", "date": "2026-08-14"}], "tags": [], "use_cases": ["local llm inference", "fast local llm inference", "local inference on rtx 5090"], "compared": [{"slug": "llama-cpp", "name": "llama.cpp", "co": 1, "switched_from": 1, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49554520", "title": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49432317", "title": "Qwen 3.8-Flash-Next releasing tomorrow (125B a6B)", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49299605", "title": "Qwen 3.8 27B is out: open weights, best local dense model yet", "mentions": 1}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {}, "2026-04": {}, "2026-05": {}, "2026-06": {}, "2026-07": {}, "2026-08": {"recommend": 2}, "2026-09": {"recommend": 1}, "2026-10": {}}, "quotes": [{"id": 792884672, "item_id": 49555292, "url": "https://news.ycombinator.com/item?id=49555292", "by": "pllbnk", "date": "2026-09-03", "stance": "recommend", "quote": "on RTX 5090 I can now get ~200 tok/s and over 400 tok/s on concurrent requests which is plenty fast for a local model", "excerpt": "Just a couple days ago I learned about ninfer (https://github.com/Neroued/ninfer) and on RTX 5090 I can now get ~200 tok/s and over 400 tok/s on concurrent requests which is plenty fast for a local model of this strength.", "context": "", "use_case": "local LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 790958528, "item_id": 49434908, "url": "https://news.ycombinator.com/item?id=49434908", "by": "Philpax", "date": "2026-08-25", "stance": "recommend", "quote": "Strongly recommend https://github.com/Neroued/ninfer, which can pull ~180 TPS on 5090 with 3.8", "excerpt": "Strongly recommend https://github.com/Neroued/ninfer, which can pull ~180 TPS on 5090 with 3.8, and 500 (!) with 3.6 35B-A3B.", "context": "runs on an RTX 5090", "use_case": "fast local LLM inference", "switched_from": "", "story": "Qwen 3.8-Flash-Next releasing tomorrow (125B a6B)"}, {"id": 788895008, "item_id": 49305938, "url": "https://news.ycombinator.com/item?id=49305938", "by": "kimsey0", "date": "2026-08-14", "stance": "recommend", "quote": "as inference engine gets me ~138 tokens/second, roughly double what I get with a naive llama.cpp setup.", "excerpt": "If anyone else is running this on an RTX 5090, https://github.com/Neroued/ninfer as inference engine gets me ~138 tokens/second, roughly double what I get with a naive llama.cpp setup.", "context": "runs an RTX 5090", "use_case": "local inference on RTX 5090", "switched_from": "llama.cpp", "story": "Qwen 3.8 27B is out: open weights, best local dense model yet"}]}