# Muse Spark 161 Hacker News commenters recommended Muse Spark and 142 criticized it, November 2025 to November 2026. It was recommended in 7 of 13 months. **Verdict:** Meta's model is recommended as cheap and close to frontier models, with versions that run on consumer hardware, though many find it behind Claude for coding. ## Why people recommend it - Competitive with frontier models (9 comments): "If it slightly beats or even matches Opus 4.6, it means Meta is capable of building a model competitive with the leading AI company." ([source](https://news.ycombinator.com/item?id=47693187)) "I tried multiple riddles, graphs and questions I know some LLMs fails at, but this one seems to do well." ([source](https://news.ycombinator.com/item?id=47705615)) "Sure it's not as good as Claude right now but for their first model in years it's certainly not bad." ([source](https://news.ycombinator.com/item?id=47695713)) - Cheap, good value (6 comments): "It seems to trade blows with GPT 5.5 and Opus 4.8 in performance while being cheaper than GLM 5.2." ([source](https://news.ycombinator.com/item?id=48847395)) "if it actually is this close to opus, but at a third the cost, that's pretty solid" ([source](https://news.ycombinator.com/item?id=49189046)) "combined with almost free model (muse spark contributor) it's actually pretty good" ([source](https://news.ycombinator.com/item?id=49192088)) - Runs locally on consumer hardware (5 comments): "optimized to run on Mac Studio with M5 Ultra, and together they're the best way to run a powerful model on your own desktop" ([source](https://news.ycombinator.com/item?id=50107287)) "They've done considerable effort in making sure this model can run on basically every model runner" ([source](https://news.ycombinator.com/item?id=50107063)) "this is also the first time I've seen an actual lab announce they optimized for consumer hardware" ([source](https://news.ycombinator.com/item?id=50109940)) - Capable for coding work (2 comments): "perfectly capable of following my spec and then poking holes in the implementation till there are none left" ([source](https://news.ycombinator.com/item?id=49811750)) "This is already so good that 90% of my work has shifted away from codex to this." ([source](https://news.ycombinator.com/item?id=50059995)) ## What people warn about - Lags Claude/Opus for programming (5 comments): "It's a decent model if the benchmarks are to be believed, but it won't be close to Opus in usefulness for programming." ([source](https://news.ycombinator.com/item?id=47695003)) "Sure it's not as good as Claude right now but for their first model in years it's certainly not bad." ([source](https://news.ycombinator.com/item?id=47695713)) "It's looking rather low on reasoning and long-range problems with the approach described." ([source](https://news.ycombinator.com/item?id=47696437)) - Coding quirks: stuck loops, assumptions (4 comments): "can get stuck outputting similar sounding thought summaries with no apparent progress when asked to solve bugs" ([source](https://news.ycombinator.com/item?id=49541545)) "MS1.2 wrote C code without following original source code writing style, and no descriptive info why writing such code" ([source](https://news.ycombinator.com/item?id=49546448)) "Gemini 3.1 pro or Muse Spark 1.2 or the new Claude 5 models just assume so much without making a quick check in" ([source](https://news.ycombinator.com/item?id=49309625)) - Not truly open; restrictive license (4 comments): "It's a bit of “Open”-washing but reasonable to the model trainer for their investment" ([source](https://news.ycombinator.com/item?id=50108781)) "The model license requires large inference providers to contact Meta separately to obtain a license to serve the model." ([source](https://news.ycombinator.com/item?id=50107323)) "It's not open, like the Llama. There is a licensing agreement by which they allow you to use it for free on conditions." ([source](https://news.ycombinator.com/item?id=50107910)) - Distrust of Meta and benchmarks (5 comments): "there is no way meta is not training on your data regardless of your monthly invoice" ([source](https://news.ycombinator.com/item?id=49190647)) "I don't see a reason to use something that is, at best, about as good as models that are produced by labs who don't have a reputation of cheating." ([source](https://news.ycombinator.com/item?id=48861823)) "this looks like something they rolled out in a hurry to take advantage of the FOMO of the uninformed" ([source](https://news.ycombinator.com/item?id=50107867)) ## Alternatives hackers recommend - [Claude Opus](https://hackerapproved.com/p/claude-opus.md): 3331 recommend - [OpenAI GPT](https://hackerapproved.com/p/gpt-5-5.md): 1205 recommend - [Claude Sonnet](https://hackerapproved.com/p/claude-sonnet.md): 854 recommend - [DeepSeek V4 Flash](https://hackerapproved.com/p/deepseek-v4-flash.md): 828 recommend - [Fable](https://hackerapproved.com/p/fable.md): 823 recommend ## Where to get it - [Visit the official site](https://www.meta.ai/) ## Cite as Hacker Approved, "Muse Spark", https://hackerapproved.com/p/muse-spark/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).