Muse Spark 1.2

Why teams choose it

Tradeoffs to know

• Muse Spark 1.2 weights are not part of the Muse Glimmer open-weight release; use the hosted API unless Meta publishes a separate checkpoint.
• The low-cost contributor tier permits use of prompts and completions to improve Meta products; review data policy before sending proprietary code.
• Vendor evaluations are harness-sensitive and should be reproduced with the intended tools, approvals, and repository tests.

When not to use this

• Not ideal for simple tasks where cheaper models in the same lineup are good enough.
• Avoid for regulated or high-stakes outputs without evaluations that mimic your tooling, data, and review process.
• Pair catalog notes with comparisons and your own benchmarks before declaring a routing winner.

Technical specs

Inputs

text, image, video, PDF

Outputs

text

Capabilities

coding, complex debugging, codebase understanding, long-horizon agents, tool calling, planning, context compaction, multi-agent workflows

License

Proprietary API

Model string

muse-spark-1-2

Benchmarks

{
  "source": "https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2",
  "vendor_reported": true
}