Meta5 Aug 2026Model

Muse Spark 1.2

Benchmark results

Benchmark	Score	Source check
AutomationBenchxhigh effort; public v3 task set; pass@1 · reported later, in Introducing Muse Spark 1.3	38.2%	Checked A person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
DeepSWE v1.1Muse Code harness; xhigh effort; no internet; averaged over 5 attempts	59.3%	Checked A person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
Datacurve leaderboard; xhigh effort; harness not stated · reported later, in Introducing Muse Spark 1.3	55.0%	Checked A person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
Terminal-Bench 2.1Muse Code harness; xhigh effort; averaged over 5 attempts	82.9%	Corroborated The same number is printed in 1 other lab document: 1
MCP Atlasrun by Scale AI with the benchmark's own harness	90.3%	Checked A person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.

Sources

Reported by Meta