# Scale Agentic Coding Workflows with Meta’s Muse Spark Open on NVIDIA GB300 NVL72 > Source: > Published: 2026-10-15 18:00:01+00:00 Meta launches Muse Spark Open, a coding-optimized 816B [mixture-of-experts (MoE)](https://www.nvidia.com/en-us/glossary/mixture-of-experts/) model with 42B active parameters and a 1M token context window. Trained on [NVIDIA GB300 NVL72 Blackwell Ultra](https://www.nvidia.com/en-us/data-center/gb300-nvl72/) specifically for agentic coding workflows, it delivers higher first attempt accuracy, reliable tool calling, and deep codebase understanding across repositories. The 1M token context window holds an entire project, including complex dependencies, legacy code, and thousands of files without breaking a task into pieces.   As a reasoning model, Muse Spark Open adds audio and video perception alongside image and document understanding, working natively and reliably across what it sees, hears, and reads. [NVFP4](https://developer.nvidia.com/blog/introducing-nvfp4-for-efficient-and-accurate-low-precision-inference/) checkpoints generated by Meta and optimized for NVIDIA Blackwell Ultra shrink the memory footprint of all 816B parameters with native FP4 Tensor Core acceleration. Fewer bytes moved per token to increase tokens per second per user, and the freed-up memory supports larger KV cache for concurrent long context sessions, delivering higher throughput, a lower cost per token.  ## Custom-built for agentic coding multimodal understanding across text, image, video, and audio  General coding models are optimized for single-shot completions such as generating a function, writing a unit test, or suggesting an inline fix. They excel at discrete, well-scoped tasks where the context is small, and the answer is immediate. Agentic coding demands something fundamentally different. Planning a multifile refactor, navigating thousands of files, recovering from failing tests, and coordinating subagents over long horizons requires a model built for sustained, multistep work. Muse Spark Open is built for exactly these workloads, delivering higher first-attempt accuracy, reliable tool calling, and deep codebase understanding across repositories. It plans, acts, observes, and self corrects. It supports multistep task execution with tool use, output evaluation, and iterative refinement. The model uses goal conditioning to hold direction and context compaction to retain the knowledge needed to sustain progress across long-running sessions. With a 1M token context window, it holds an entire project, including complex dependencies, legacy code, and thousands of files without breaking tasks into pieces. Beyond coding, it searches the web (including multimodal browsing), spawns and coordinates subagents for decomposable tasks, and leverages computer use capabilities unified across desktop, mobile, and browser. It generates separate commentary before each action, so developers know what the agent is doing and why. The model’s 1.8B perception encoder natively processes interleaved text, images, video frames, and audio without separate pipelines per modality. This means that the model reasons over what it sees, hears, and reads in a single unified pass. This suits developer workflows spanning screenshots, diagrams, error logs, and video walkthroughs within the same agent session. Sparse MoE routing activates only 42B of 816B total parameters per token, delivering frontier agentic capability at a fraction of the serving cost of dense models at equivalent scale. ## Muse Spark Open optimized performance on GB300 NVL72  The [NVIDIA GB300 NVL72](https://www.nvidia.com/en-us/data-center/gb300-nvl72/) features a rack scale architecture that integrates 72 NVIDIA Blackwell Ultra GPUs into a single platform. Its large, 72-GPU [NVIDIA NVLink](https://www.nvidia.com/en-us/data-center/nvlink/) domain enables efficient communication at 130 TB/s, eliminating bottlenecks that appear when expert traffic must cross traditional off-the-shelf networks. For a sparse MoE model like Muse Spark Open, which routes each token to eight of 256 experts, that bandwidth directly determines how much of the theoretical throughput a deployment realizes.  Agentic workloads turn static single-turn interactions into multistep workflows that reason, invoke tools, coordinate subagents, and carry growing context from one turn to the next. Out of the box, Muse Spark Open running agentic workloads on NVIDIA GB300 NVL72 delivers up to 50K per second per GPU and up to 600 tokens per second per user, enabling AI factories to run large parameter models in production at high throughput and low latency. ## Flexible deployment options for Muse Spark Open  NVIDIA supports multiple inference stacks to meet a variety of developer needs. [SGLang,](https://docs.sglang.io/cookbook/autoregressive/Meta/MuseSparkOpen) [vLLM](https://recipes.vllm.ai/meta-models/Muse-Spark-Open-816B-A42B), and [TensorRT LLM](https://github.com/NVIDIA/TensorRT-LLM/blob/feat/musespark-open/docs/source/deployment-guide/deployment-guide-for-musespark-open-on-trtllm.md) give developers direct control over serving configuration, parallelism, and batching on the NVIDIA accelerated computing platform. [TensorRT LLM](https://github.com/NVIDIA/TensorRT-LLM/blob/feat/musespark-open/docs/source/deployment-guide/deployment-guide-for-musespark-open-on-trtllm.md) adds compiled engines tuned for NVIDIA GPUs. NVIDIA Dynamo publishes deployment recipes for all three~~,~~ and separates prefill from decode across dedicated GPU pools, so compute bound prompt processing doesn’t slow token generation. Its KV cache-aware routing sends each request to the worker already holding the relevant prefix, avoiding precomputation as context accumulates across long agentic sessions.  ## Post-train Muse Spark Open with NeMo AutoModel and Megatron Bridge  Developers can post-train Muse Spark Open on domain specific datasets using [NVIDIA NeMo AutoModel](https://github.com/NVIDIA-NeMo/Automodel/blob/main/docs/model-coverage/vlm/meta-models/muse_spark_open.mdx), a PyTorch native training library with native Hugging Face checkpoint support and no model conversion required. It enables full SFT and LoRA fine tuning out of the box, optimized for rapid experimentation over a wide variety of industry specific cases.  When those workloads outgrow a single node, [NVIDIA Megatron Bridge](https://github.com/NVIDIA-NeMo/Megatron-Bridge/tree/main/examples/models/muse_spark_open) takes the same Hugging Face checkpoints and adds Megatron Core tensor, pipeline, and expert parallelism, so teams scale from experimentation to production training without reconverting the model.  ## Get started with GPU-accelerated Muse Spark Open  To get started, [download Muse Spark Open NVFP4 weights](https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B-NVFP4) from Hugging Face. For production deployments, [try the downloadable NVIDIA NIM](https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B-NVFP4). The NIM reduces setup to a single command, replacing benchmarking and tuning cycles with a container that automatically selects the best serving backend and precision for the detected hardware.