‹ PublicAI Index
The LLM benchmark aggregator.
Inkling 2 Small
Thinking Machines
Strongest in Legal (#11 of 173), weakest in Human preference (#121 of 354). Above par in 14 of 15 scopes. Among the models it meets almost everywhere, it finishes behind Gemini 4 Carbon and Muse Wonder Full and ahead of Gemini 3.7 Flash and Gemini 3.5 Flash.
§ 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.
Agents59.4−7#18/2982/6
Knowledge work65.1−5.6#21/2082/2
Coding55−11.2#36/1791/6
Agentic coding55.5−11.4#35/1731/5
Reasoning54.7−10.7#50/1951/5
Reasoning58.4−10.6#29/1471/3
Professional53.1−9.3#58/1941/2
Legal60.7−5.2#11/1731/1
Biology research50.6−20.1#15/311/1
Cybersecurity57−6.9#16/471/1
IT operations40.5−33.3#26/331/1
Core abilities51.6−13.7#81/1991/4
General intelligence52.4−17#68/1781/3
Human preference57.8−11#121/3541/1
Human preference57.8−11#121/3541/1
§ 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 Inkling 2 Small, left for the other.
§ 3 · Sources
Where the numbers come from
6 publications, 11 figures. Every one links to the page it was read from.
LMArena Text 1412
Artificial Analysis Intelligence Index 31
GDPval-AA 1630
AA-Briefcase 1527
ARC-AGI-2 81.7%
Vals · Legal Research Bench 46.63%Vals · Harvey Legal Agent Benchmark 21.67%Vals · BioMysteryBench 70%Vals · CyberBench 72.74%Vals · SRE Bench 1.91%Vals · Vibe Code Bench 81.55% (OpenHands)
Badge
[](https://publicai.io/model-index/m/inkling-2-small)