Meta: Muse Spark 1.2

Meta: Muse Spark 1.2 is a model from Meta that accepts text, image, video, and file inputs. It offers a context window of 1,048,576 tokens and a maximum completion length of 943,718 tokens, so long documents, large codebases, or extended media inputs can fit in a single request, with room for very long outputs. It supports tool use and reasoning. The available facts do not confirm structured output support or whether the weights are open. Pricing is $1.25 per million input tokens and $4.25 per million output tokens. That suits teams that send large inputs or need long generations and want to keep per-token costs in view. It belongs on a shortlist for workloads that mix video, images, and files with tool calls across very long contexts. There is no independent benchmark coverage yet, so its quality is unproven; test it on your own tasks before committing.