‹ PublicAI Index
The LLM benchmark aggregator.
Falcon H1 0.5B Instruct
TII · 0.5B
Strongest in Fairness (#161 of 299, on 1 of its 2 boards), weakest in Safety (#293 of 336). Above par in 0 of 6 scopes. Among the models it meets almost everywhere, it finishes behind Claude Sonnet 5 and Claude Opus 5.5 and ahead of Aya 23 8B and Grok 3.
§ 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.
Safety44.4−17.1#293/3361/3
Fairness46.4−23.6#161/2991/2
Jailbreak resistance47.8−17.1#193/2721/1
Secure code41.1−25.4#210/2741/1
Toxicity avoidance47.7−11.3#216/2721/1
Harm refusal41.2−20.4#285/2991/2
§ 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 Falcon H1 0.5B Instruct, left for the other.
§ 3 · Sources
Where the numbers come from
1 publication, 6 figures. Every one links to the page it was read from.
Enkrypt · Jailbreak risk 14.6%Enkrypt · Harmful content risk 65.6%Enkrypt · CBRN risk 20.8%Enkrypt · Toxicity risk 7.9%Enkrypt · Bias risk 83.5%Enkrypt · Insecure code risk 51.6%
Badge
[](https://publicai.io/model-index/m/falcon-h1-0-5b-instruct)