Muse Spark Open is an open weights reasoning model from Meta for coding and agentic workflows. It supports tool calling and native image understanding in a single checkpoint. It uses a sparse Mixture-of-Experts architecture with 816B total / 42B active parameters, routing each token to 8 of 256 experts, and hybrid attention that repeats three sliding-window layers (2,048-token window) for every full-attention layer. Inputs: text and images. Output: text.
→ Limited preview: Available on serverless as a limited preview through October 29, 2026, on US-only capacity, with the usual 50% US-only premium waived.
→ Production: For dedicated capacity, deploy on-demand or move to reserved throughput.
→ Fine-tuning: Train LoRA adapters through the Serverless Training API using base model accounts/fireworks/models/muse-spark-open-nvfp4. Managed fine-tuning in the Fireworks UI isn’t available for this model yet. Fine-tuned models deploy on-demand.
Muse Spark Open is available via Fireworks' serverless API, where you pay per token. There are several ways to call the Fireworks API, including Fireworks' Python client, the REST API, or OpenAI's Python client.
On-demand deployments allow you to use Muse Spark Open on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits.
Run queries immediately, pay only for usage
