{"slug": "muse-spark-open", "name": "Muse Spark Open", "category": "ai-models", "category_name": "AI models", "description": "", "recommenders": 4, "commenters": 7, "warners": 0, "score": 4.853, "rising": 4.0, "first": "2026-10-15", "latest": "2026-10-16", "why": "one of the best open models I've seen for conversion, which is impressive.", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 4, "neutral": 3, "compare": 0, "warn": 0}, "spark": [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4], "listed": true, "loved": [], "caveat": "", "short_quote": "one of the best open models I've seen for conversion, which is impressive.", "kind": "", "icon": false, "family": "muse", "domain": "artificialanalysis.ai", "github_repo": null, "official_url": "", "pricing": "", "aliases": ["Muse Spark Open"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 801951456, "item_id": 50121966, "url": "https://news.ycombinator.com/item?id=50121966", "by": "scirob", "date": "2026-10-16", "version": null, "stance": "recommend", "quote": "So for German Language this is now actually the #1", "excerpt": "So for German Language this is now actually the #1 model from what I can see https://dach.peerbench.ai/compare?models=meta%2Fmuse-spark-open,z-ai%2Fglm-5.3-flash,qwen%2Fqwen3.8-flash So it does mean something to be the top western model if it means better performance for western languages. sry late benchmarks failed first few runs.", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801938064, "item_id": 50121129, "url": "https://news.ycombinator.com/item?id=50121129", "by": "wolttam", "date": "2026-10-16", "version": null, "stance": "recommend", "quote": "This is still a welcome model, even if I won't be running it at home", "excerpt": "Most people are probably looking at: https://artificialanalysis.ai/models/comparisons?compare=muse-spark-open,glm-5-3-flash,qwen3-8-flash-next This is still a welcome model, even if I won't be running it at home", "context": "", "use_case": "Open-weights model", "switched_from": "", "story": "Apple ML Research: Muse Spark Open with MLX on Mac Studio"}, {"id": 801799009, "item_id": 50112438, "url": "https://news.ycombinator.com/item?id=50112438", "by": "GeekyBear", "date": "2026-10-16", "version": null, "stance": "recommend", "quote": "we show how to run the latest frontier-class model \u2014 Muse Spark Open", "excerpt": "Apple research has also posted information on the topic. > In this post, we show how to run the latest frontier-class model \u2014 Muse Spark Open, with nearly a trillion parameters \u2014 right from your desk, using the latest version of MLX, released today. On a cluster of four Mac Studios, Muse Spark Open generates over 36 tokens per second Beginning on a single Mac Studio (up to 512GB of unified memory), then scaling out a", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801740608, "item_id": 50108788, "url": "https://news.ycombinator.com/item?id=50108788", "by": "jjcm", "date": "2026-10-15", "version": null, "stance": "recommend", "quote": "one of the best open models I've seen for conversion, which is impressive.", "excerpt": "Image->html test for muse spark open. Original designs: https://diffui.ai/app/canvas/5093e689-1e74-4f26-b632-2a4500f0da97?share=65ac3348-895b-4264-824e-b5f343040009 Muse Spark Open: https://html.non.io/lcars-muse-spark/ Opus 5.5 (sota): https://html.non.io/lcars-opus-5.5/ Overall it actually does pretty solid. It doesn't completely follow the design, but it gets the vibe. It's one of the best open models I've seen fo", "context": "", "use_case": "Image to HTML conversion", "switched_from": "", "story": "Muse Spark Open"}]}], "caveats": [], "who": [], "tags": [], "use_cases": ["open-weights model", "image to html conversion"], "compared": [{"slug": "mlx", "name": "MLX", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=50107379", "title": "Muse Spark Open", "mentions": 6}, {"url": "https://news.ycombinator.com/item?id=50118320", "title": "Apple ML Research: Muse Spark Open with MLX on Mac Studio", "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": {}, "2026-09": {}, "2026-10": {"neutral": 3, "recommend": 4}}, "versions": [], "version_unnamed": 7, "family_ref": {"slug": "muse", "name": "Muse"}, "runs_on": null, "lines": [], "quotes": [{"id": 801951456, "item_id": 50121966, "url": "https://news.ycombinator.com/item?id=50121966", "by": "scirob", "date": "2026-10-16", "version": null, "stance": "recommend", "quote": "So for German Language this is now actually the #1", "excerpt": "So for German Language this is now actually the #1 model from what I can see https://dach.peerbench.ai/compare?models=meta%2Fmuse-spark-open,z-ai%2Fglm-5.3-flash,qwen%2Fqwen3.8-flash So it does mean something to be the top western model if it means better performance for western languages. sry late benchmarks failed first few runs.", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801938064, "item_id": 50121129, "url": "https://news.ycombinator.com/item?id=50121129", "by": "wolttam", "date": "2026-10-16", "version": null, "stance": "recommend", "quote": "This is still a welcome model, even if I won't be running it at home", "excerpt": "Most people are probably looking at: https://artificialanalysis.ai/models/comparisons?compare=muse-spark-open,glm-5-3-flash,qwen3-8-flash-next This is still a welcome model, even if I won't be running it at home", "context": "", "use_case": "Open-weights model", "switched_from": "", "story": "Apple ML Research: Muse Spark Open with MLX on Mac Studio"}, {"id": 801799009, "item_id": 50112438, "url": "https://news.ycombinator.com/item?id=50112438", "by": "GeekyBear", "date": "2026-10-16", "version": null, "stance": "recommend", "quote": "we show how to run the latest frontier-class model \u2014 Muse Spark Open", "excerpt": "Apple research has also posted information on the topic. > In this post, we show how to run the latest frontier-class model \u2014 Muse Spark Open, with nearly a trillion parameters \u2014 right from your desk, using the latest version of MLX, released today. On a cluster of four Mac Studios, Muse Spark Open generates over 36 tokens per second Beginning on a single Mac Studio (up to 512GB of unified memory), then scaling out a", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801740608, "item_id": 50108788, "url": "https://news.ycombinator.com/item?id=50108788", "by": "jjcm", "date": "2026-10-15", "version": null, "stance": "recommend", "quote": "one of the best open models I've seen for conversion, which is impressive.", "excerpt": "Image->html test for muse spark open. Original designs: https://diffui.ai/app/canvas/5093e689-1e74-4f26-b632-2a4500f0da97?share=65ac3348-895b-4264-824e-b5f343040009 Muse Spark Open: https://html.non.io/lcars-muse-spark/ Opus 5.5 (sota): https://html.non.io/lcars-opus-5.5/ Overall it actually does pretty solid. It doesn't completely follow the design, but it gets the vibe. It's one of the best open models I've seen fo", "context": "", "use_case": "Image to HTML conversion", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801730928, "item_id": 50108183, "url": "https://news.ycombinator.com/item?id=50108183", "by": "GodelNumbering", "date": "2026-10-15", "version": null, "stance": "neutral", "quote": "39 on AA[1], about the same as Deepseek v4.1 flash and Mimo 2.5 Flash.", "excerpt": "39 on AA[1], about the same as Deepseek v4.1 flash and Mimo 2.5 Flash. Behind Glm 5.3 flash and Qwen3.8-Flash-Next [2] [1] https://artificialanalysis.ai/models/muse-spark-open [2] https://artificialanalysis.ai/models/open-source?models=mimo-v2-6-pro%2Cglm-5-3%2Ckimi-k3%2Cglm-5-3-flash%2Cdeepseek-v4-1-flash%2Cmuse-spark-open%2Cqwen3-8-27b%2Ck2-horizon-375b-a23b%2Cminimax-m3%2Cnemotron-3-5-super-vl%2Cinkling%2Cmuse-gli", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801725568, "item_id": 50107848, "url": "https://news.ycombinator.com/item?id=50107848", "by": "spwa4", "date": "2026-10-15", "version": null, "stance": "neutral", "quote": "You & Me won't be running this one.", "excerpt": "https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B-NVFP4 816B model, A42B active parameters Unsloth and Meta have guides on how to run the model, but allow me to summarize: You & Me won't be running this one. https://unsloth.ai/docs/models/muse-spark-open", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}, {"id": 801713008, "item_id": 50107063, "url": "https://news.ycombinator.com/item?id=50107063", "by": "armcat", "date": "2026-10-15", "version": null, "stance": "neutral", "quote": "They've done considerable effort in making sure this model can run on basically every model runner", "excerpt": "They've done considerable effort in making sure this model can run on basically every model runner, from Unsloth to MLX, https://dev.meta.ai/docs/muse-spark-open/deploy", "context": "", "use_case": "", "switched_from": "", "story": "Muse Spark Open"}]}