Muse Spark 1.3 Contributor

Meta's cost-efficient tier of its multimodal reasoning model, designed for experimentation and early-stage agentic workflows with a 1M-token context window.

Strengths

Handles extended inputs and multimodal reasoning tasks across text, images, and tool use at a lower cost than the standard Muse Spark 1.3 tier.

Best for

Experimentation, learning, and early-stage agentic or multi-agent workflows that benefit from a large context window.

Limitations

A newer version of an earlier contributor tier; production workloads requiring higher quality reasoning should use the standard Muse Spark 1.3.