# Can you run Muse Spark Open locally? Hardware, RAM and cost (slim) > Vlad Cealicu on Meta's local pitch for Muse Spark Open (816B, 42B active), and how to run it locally. How much RAM and VRAM each version needs, whether it runs on a Mac Studio, a Minisforum MS-S1 Max, RTX 5090s, a DGX Spark or a CPU server, how many of each you'd need, what they cost and how fast they'd be, and when an API is cheaper. - Full: https://www.anchorterminal.com/blog/run-muse-spark-open-locally.md (~4,250 tokens) · this version ~2,130 tokens · JSON https://www.anchorterminal.com/blog/run-muse-spark-open-locally.json · canonical https://www.anchorterminal.com/blog/run-muse-spark-open-locally - Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-15 - Licence: CC BY 4.0, credit Anchor Terminal (anchorterminal.com) and link this page when you pass anything on By Vlad Cealicu, 15 October 2026. 2,015 words, about 9 min to read. Thank you, Meta, for opening the weights of [Muse Spark Open](/tools/muse-spark-open), released today, with 816B parameters, 42B of them active for each token, a 1M-token context, and it takes text, image, video and audio [[1]](#ref-1). Meta's post leads with running it on your desk, on two Mac Studios with M5 Ultra and 512GB of unified memory each. ## Can most people run Muse Spark Open locally? For the labs and the big companies running AI hardware, fine. Below is how much memory each version needs, every way I could find to hold it, what each costs on 15 October 2026 and how fast it would be. ## How much RAM does Muse Spark Open need? It depends on which version of the weights you run. Weights on disk, in GB | Checkpoint | GB | | --- | --- | | BF16, Meta's full precision | 1640 | | Q8_0 (ggml-org) | 872 | | NVFP4 (Meta, what the Macs run) | 537 | | UD-Q4_K_XL (Unsloth) | 506 | | Q4_K_S (ggml-org) | 470 | | UD-Q2_K_XL (Unsloth, its suggested start) | 311 | | UD-IQ1_S (Unsloth, the smallest) | 176 | Each series has a zero-based scale. On top of the file you need memory for the context and the runtime. ## Can I run Muse Spark Open on a Mac Studio? Yes, on two or four of them, the way Meta does it with MLX [[4]](#ref-4). Two Mac Studios with M5 Ultra and 512GB each hold NVFP4 with room to spare, but the 512GB model isn't orderable until late October and Apple hasn't announced its price [[6]](#ref-6). One 512GB Mac Studio, once it's on sale, can hold Unsloth's 2-bit file (311GB) or anything smaller, but not NVFP4. ## Can I run Muse Spark Open on a Minisforum MS-S1 Max or another Strix Halo mini PC? Yes, several of them, with llama.cpp splitting the model across the network. AMD has published the recipe for a model this size on four Ryzen AI Max+ 395 machines with 128GB each, giving 120GB of each to the GPU and linking them with llama.cpp over Ethernet [[14]](#ref-14). AMD measured 9.45 tokens a second generating with Kimi K2.5 at 2-bit (375GB) on four machines over 5Gbps Ethernet [[14]](#ref-14). ## Can I run Muse Spark Open on an RTX 5090 or other gaming GPUs? Technically yes, but it doesn't make sense. For the same money, eight RTX PRO 6000 Blackwell cards with 96GB each hold the 4-bit NVFP4 in 768GB of VRAM, at $8,900 to $11,500 a card, so $71,000 to $92,000 before the server [[11]](#ref-11). The setup people with one or two gaming GPUs use for models like this is llama.cpp with the experts kept in system RAM and the rest on the GPU. ## Can I run Muse Spark Open on a DGX Spark or a DGX Station? A DGX Spark has 128GB and costs $6,950 for the 128GB model [[17]](#ref-17), so the sums are the same as for the mini PCs. A DGX Station GB300 has 748GB, $92,887 from Supermicro, but only 252GB of it is HBM on the GPU and the other 496GB is system memory [[9]](#ref-9). ## Can I run Muse Spark Open on a CPU server? Yes, slowly, and it's the cheapest way to hold the 4-bit model. ## What does Meta run it on? Four B300 GPUs for NVFP4 and eight for BF16 [[3]](#ref-3). ## How fast is Muse Spark Open on local hardware? Nobody has published speeds for it yet, Meta included. - 87 tok/s: two Mac Studios, NVFP4, each reading half of every layer at 1.2TB/s - 175 tok/s: four Mac Studios, the same split four ways - 19 tok/s: an EPYC with 12 channels of DDR5-4800, Q4_K_S - 9 tok/s: an EPYC with 8 channels of DDR4-3200, Q4_K_S Ceilings from memory bandwidth alone, not measurements Real speeds will be well under these, because the Macs talk to each other over Thunderbolt on every token, and every token also reads the context cache, which grows with the conversation. ## Is it cheaper to run Muse Spark Open locally or through an API? Through an API, by a long way. Renting four B300s at $28.40 an hour, you'd need to produce about 1,860 output tokens a second around the clock to beat that price, which only makes sense with a lot of users batched together. ## What I'd do If you want to try it, call a hosted provider. We'll publish what we measure once our own hardware is up. ### How much RAM do you need to run Muse Spark Open? About 537GB for Meta's NVFP4 weights, plus room for the context. ### Can Muse Spark Open run on one Mac Studio? Not the version Meta runs. ### How many Minisforum MS-S1 Max or other 128GB Strix Halo machines do you need? Two for the 1-bit file, three for 2-bit and five for 4-bit, giving 120GB of each to the GPU and linking them with llama.cpp over the network. ### How many RTX 5090s do you need for Muse Spark Open? Six for the 1-bit file, eleven for 2-bit and sixteen for 4-bit, at 32GB each. ### Is it cheaper to run Muse Spark Open locally or through an API? Through an API. --- - Older: [Few people have seen what agents can do, so little is built for them](https://www.anchorterminal.com/blog/few-have-seen-what-agents-can-do.md) - Newer: [How to make your site and tool ready for agents](https://www.anchorterminal.com/blog/how-to-make-your-site-ready-for-agents.md) - All posts: https://www.anchorterminal.com/blog/index.md ## References Read on 15 October 2026. [1] Meta, Muse Spark Open, research.meta.ai, 15 October 2026. Source for the release, the parameter counts, the context, the modalities, the two Mac Studios with 512GB each and the "terabyte of unified memory" quote. [2] Meta, Build with Muse Spark Open, dev.meta.ai, 15 October 2026. Source for the runtimes, the checkpoints from Meta, AMD, Intel and Unsloth, and the MLX setup on two or four Mac Studios. [3] Meta, Get the model, Muse Spark Open docs. [4] Meta, Run inference, Muse Spark Open docs. [5] Hugging Face, the file lists of unsloth/Muse-Spark-Open-816B-A42B-GGUF and ggml-org/Muse-Spark-Open-816B-A42B-GGUF, 15 October 2026. Source for the file sizes in the chart. [6] Joe Rossignol, Mac Studio With M5 Ultra Chip and 512GB of RAM Launching in October, MacRumors, 25 August 2026, and the MacRumors Mac Studio roundup. [7] Macworld, Apple announces the M5 Ultra Mac Studio with up to 512GB of RAM, 25 August 2026. Source for the $5,499 starting price and the 1.2TB/s of memory bandwidth. [8] RunAIHome, Apple M5 Ultra Mac Studio for Local AI in 2026, 1 September 2026. Source for the 256GB configuration at about $9,499 ($5,499 plus Apple's $4,000 upgrade). [9] TechRadar, Nvidia GB300 DGX Station desktop PC goes on sale, 1 September 2026. Source for the $92,887 price from Supermicro and the 748GB split into 252GB of HBM3e and 496GB of system memory. [10] Thunder Compute, NVIDIA B300 pricing, updated 9 October 2026. Source for the B300 rental prices by provider and the purchase prices of a B300 and an eight-GPU HGX B300 system. [11] Digital Citizen, NVIDIA RTX PRO 6000 Blackwell prices climb past $10,000, 18 May 2026. Source for the street prices from $8,900 to $11,500. https://www.digitalcitizen.life/nvidia-rtx-pro-6000-blackwell-prices-climb-past-10000-as-ai-demand-keeps-rising/ [12] OpenClawDC, Used EPYC servers for CPU-only MoE inference in 2026, 24 August 2026. Source for the RAM prices per GB, the used CPU prices and the self-reported speeds for DeepSeek R1 on EPYC servers. [13] Anchor Terminal, the Muse Spark Open listing, 15 October 2026. Source for the hosted prices through OpenRouter, Groq's preview, the gated download and the licence. [14] AMD, How to Run a One Trillion-Parameter LLM Locally, an AMD Ryzen AI Max+ cluster guide, 25 February 2026. Source for the four-machine cluster, 120GB of GPU memory per machine, llama.cpp over 5Gbps Ethernet and 9.45 tokens a second for Kimi K2.5 at 2-bit. [15] Minisforum UK, MS-S1 Max product page, 15 October 2026. Source for the £3,519 sale price of the 128GB model with 2TB, the two 10GbE ports and the USB4 ports at 40 and 80Gbps. [16] GPU Poet, NVIDIA GeForce RTX 5090 price in October 2026. Source for the October average of $6,195 against an MSRP of $1,999. https://gpupoet.com/gpu/learn/price/october-2026/nvidia-geforce-rtx-5090 [17] KuCoin News, NVIDIA launches 64GB DGX Spark at $4,999, 128GB version at $6,950, 2 October 2026. Source for the DGX Spark prices.