‹ PublicAI Index
The LLM benchmark aggregator.
Gemma 2B It
Google · 2B
Strongest in Harm refusal (#59 of 300, on 1 of its 2 boards), weakest in Human preference (#327 of 342). Above par in 2 of 8 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Opus 4.7 and ahead of Command R and Qwen2 7B Instruct.
§ 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.
Safety43.7−17.8#298/3371/3
Harm refusal55.6−6#59/3001/2
Jailbreak resistance51.8−13.1#163/2721/1
Toxicity avoidance26−33#262/2721/1
Secure code26−40.5#273/2741/1
Fairness38.4−31.6#285/3001/2
Human preference27.6−39.9#327/3421/1
Human preference27.6−39.9#327/3421/1
§ 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 Gemma 2B It, left for the other.
§ 3 · Sources
Where the numbers come from
2 publications, 7 figures. Every one links to the page it was read from.
LMArena Text 1094
Enkrypt · Jailbreak risk 11.3%Enkrypt · Harmful content risk 22.8%Enkrypt · CBRN risk 5.7%Enkrypt · Toxicity risk 29.6%Enkrypt · Bias risk 95.9%Enkrypt · Insecure code risk 95.6%
Badge
[](https://publicai.io/model-index/m/gemma-2b-it)