Frontier Log

MetaModel

Muse Spark 1.1

Made by
Meta
Released
Announced in
Introducing Muse Spark 1.1ai.meta.com
Event
Muse Spark 1.1
API record
/api/v1/models/muse-spark-1-1

Benchmark results

BenchmarkScoreSource check
Humanity's Last ExamWith tools · xhigh effort; bash and browser tools62.1%
CorroboratedThe same number is printed in 1 other lab document: 1
SWE-bench ProScale AI leaderboard; mini-swe-agent harness; 731-task public set61.5%
CheckedA 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 v1.1xhigh effort; harness based on mini-swe-agent; no internet; averaged over 5 attempts53.3%
CorroboratedThe same number is printed in 1 other lab document: 1
mini-swe-agent harness; xhigh effort; no internet; averaged over 5 attempts · reported later, in Introducing Muse Code and Muse Spark 1.253.0%
CheckedA 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 2.1xhigh effort; bash-only harness; averaged over 5 attempts80.0%
CorroboratedThe same number is printed in 1 other lab document: 1
mini-swe-agent harness; xhigh effort; averaged over 5 attempts · reported later, in Introducing Muse Code and Muse Spark 1.276.2%
CheckedA 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.
OSWorld-Verifiedxhigh effort; 361 tasks without Google Drive ones; GUI tool only; 200-step limit; one attempt80.8%
CorroboratedThe same number is printed in 1 other lab document: 1
MCP Atlasrun by Scale AI88.1%
CheckedA 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

Events

Stories

All models