LetTokenBurnBenchHub

Model

MiniMax-M3

MiniMax · open weights

Knowledge & Reasoning

Knowledge & Reasoning

Benchmark	Score	Metric	Harness	Source	Confidence	Link
CritPt	3.7	accuracy	N/A	Independent	low	epoch.ai
GPQA Diamond	90.9	accuracy	N/A	Independent	high	epoch.ai
LexPilot	0.6	lca_macro-f1	N/A	Independent	medium	arxiv.org
WorldFact-Bench	16.8	accuracy	direct judgment	Independent	medium	arxiv.org

Coding

Benchmark	Score	Metric	Harness	Source	Confidence	Link
ArrivalBench	50.2	failure_rate	N/A	Independent	medium	arxiv.org
SciCode	47.1	score	N/A	Independent	medium	artificialanalysis.ai

Agentic / Tool Use

Agentic / Tool Use

Benchmark	Score	Metric	Harness	Source	Confidence	Link
APTInvestBench	46.2	coverage_rate	N/A	Independent	medium	arxiv.org
Autonomous Science Bench	5.6	success_rate	N/A	Independent	medium	arxiv.org
HICO-Bench	50.4	success_rate	N/A	Independent	medium	arxiv.org
HygieneRoboBench	67.0	resolved_rate	N/A	Independent	medium	arxiv.org
Legal Research Bench	29.8	accuracy	N/A	Independent	medium	arxiv.org
MS-Eval	41.0	avg.	mad-sae (spec)	Independent	medium	arxiv.org
PersonaCUABench	16.1	success_rate	N/A	Independent	medium	arxiv.org
RealCLI	22.7	score	N/A	Independent	medium	arxiv.org

Multimodal

Benchmark	Score	Metric	Harness	Source	Confidence	Link
ChemArgus	26.2	full_problem	text only	Independent	medium	arxiv.org
DSV-Mem	13.1	accuracy	N/A	Independent	medium	arxiv.org
Event-STVG	0.0	m-viou	N/A	Independent	medium	arxiv.org
EviLens	37.8	accuracy	N/A	Independent	medium	arxiv.org
PlotGround-1k	33.8	accuracy	N/A	Independent	medium	arxiv.org

Multilingual

Benchmark	Score	Metric	Harness	Source	Confidence	Link
EHR-Bench	83.6	constructed_ba	N/A	Independent	medium	arxiv.org

Reproduced by the community