Ling-3.1-flash

Ling-3.1-flash ranks first on 3 of the 22 benchmarks it has scores on.

Ant Group Released 13 Oct 2026 Family Ling 3 The lab's model card

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Ant Group Oct 2026

Open weights Mid-size

No Epoch Capabilities Index score yet

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Benchmark	Score	Rank	Source
WideSearch	83.3%	#2 of 17	Model card
MultiChallenge	69.8%	#5 of 36	Model card
Humanity's Last Exam	37.4%	#63 of 147	Model card

Benchmark	Score		Rank	Source
CritPt	16.3%		#8 of 14	Model card
SuperMatBench	51.2%		#1 of 1	Model card

Benchmark	Score	Rank	Source
SWE-bench Multilingual	88.5%	#5 of 70	Model card
SWE-Bench Pro	68.7%	#8 of 77	Model card
FrontierSWE	75.2%	#1 of 9	Model card
SWE Atlas	64.5%	#2 of 11	Model card
DeepSWE	59.7%	#35 of 51	Model card

Benchmark	Score	Rank	Source
τ³-bench Banking	47.8%	#1 of 20	Model card
BrowseComp	91.7%	#4 of 77	Model card
SkillsBench	68.7%	#2 of 13	Model card
AutomationBench	52.5%	#8 of 41	Model card
MCP Atlas	83%	#10 of 40	Model card
Terminal-Bench 2.1	81.2%	#24 of 53	Model card
DRACO	85.5%	#5 of 11	Model card
Terminal-Bench 4.0	46%	#11 of 23	Model card

Benchmark	Score	Rank	Source
FinSearchComp	81.7%	#1 of 7	Model card
HealthBench Professional	65.3%	#5 of 25	Model card
Vals Finance Agent v2	57.9%	#5 of 11	Model card
GDPval-AA v2.1	1,673	#5 of 8	Model card

Benchmark	Score		Rank	Source
CyberGym	87.9%		#5 of 33	Model card

The benchmarks the lab chose to headline, with the scores it gave.