‹ PublicAI Index
The LLM benchmark aggregator.
Qwen3 Next 80B A3B Thinking
Alibaba · 80B · A3B
Strongest in Safe-prompt compliance (#9 of 82), weakest in Agents (#232 of 298). Above par in 9 of 11 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 4.8 and Hy3 and ahead of Qwen3.8 Max and MiMo V2.6 Pro.
§ 1 · Profile
What it is good at
Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.
Core abilities54.8−10.8#47/2001/4
General intelligence56.9−12.1#43/1801/3
Safety54.5−6.9#60/3272/3
Safe-prompt compliance59.3−1.1#9/821/1
Factual grounding51.6−19.1#46/1041/1
Harm refusal55.4−6.1#64/2901/2
Fairness52.5−17.5#90/2901/2
Human preference53.7−15.1#160/3541/1
Human preference53.7−15.1#160/3541/1
Agents44.3−22#232/2981/6
Knowledge work38.5−32.2#162/2081/2
§ 2 · Head to head
What it beats, and what beats it
The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Qwen3 Next 80B A3B Thinking, left for the other.
§ 3 · Sources
Where the numbers come from
5 publications, 9 figures. Every one links to the page it was read from.
LMArena Text 1369
GDPval-AA 136
Kagi LLM Benchmark 66.7%
HELM Safety · HarmBench 84.6%HELM Safety · SimpleSafetyTests 99.5%HELM Safety · Anthropic Red Team 99.9%HELM Safety · BBQ 91.7%HELM Safety · XSTest 98.3%
Vectara · Factual consistency 90.7%
Badge
[](https://publicai.io/model-index/m/qwen3-next-80b-a3b-thinking)