Original sourceOfficeChaiAdditional: Let's Data Science
Summary
Meta released Muse Spark 1.3 on September 2, featuring improvements in agentic tasks and coding, available via Muse Code and Meta Model API. It scores 75.4 on DeepSWE v1.1 and 88.8 on Terminal-Bench 2.1, on par with GPT-5.6 Sol and surpassing Opus 5; it also leads GPT-5.6 Sol in OSWorld 2.0 with 66…
Key points
- Developers can compare Muse Spark 1.3 with rival models in coding and agentic tasks to inform tool selection.
- Strong performance in agentic and coding benchmarks signals Meta's growing competitiveness in open-source models.
- Developers using Muse Spark 1.3 benefit from its long-horizon task handling and tool use, but should note the absence of maximal reasoning mode.
Editorial note
This page is Code & Chain's editorial summary of public sources. It may be prepared with AI assistance and published through an automated workflow. Refer to the original sources; this content is not investment, legal, or tax advice.