‹ PublicAI Index
The LLM benchmark aggregator.
Jamba 1.6 Mini
AI21
Strongest in Science (#121 of 122), weakest in Safety (#271 of 329). Above par in 1 of 15 scopes. Among the models it meets almost everywhere, it finishes behind GPT-5.5 and GPT-5.4 and ahead of Jamba 1.5 Mini.
§ 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.
Knowledge40−22.5#183/1971/3
Academic knowledge26−37.9#123/1231/1
Professional36.9−25.5#188/1941/2
Legal43.5−22.4#133/1731/1
Medical33.5−33#140/1491/1
Finance30−32.4#159/1621/1
Reasoning33.6−31.8#194/1951/5
Science26.1−37.4#121/1221/1
Mathematics32.9−30.9#149/1521/2
Safety45.7−15.7#271/3291/3
Jailbreak resistance52.8−12.2#145/2681/1
Harm refusal46.6−14.9#208/2921/2
Fairness43.9−26.1#217/2921/2
Toxicity avoidance46.2−12.8#223/2681/1
Secure code36.6−29.6#227/2701/1
§ 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 Jamba 1.6 Mini, left for the other.
§ 3 · Sources
Where the numbers come from
2 publications, 13 figures. Every one links to the page it was read from.
Vals · LegalBench 69.73%Vals · CorpFin 38.03%Vals · TaxEval 44.6%Vals · MedQA 52.52%Vals · GPQA Diamond 32.07%Vals · MMLU Pro 30.28%Vals · AIME 0.42%
Enkrypt · Jailbreak risk 10.5%Enkrypt · Harmful content risk 61.7%Enkrypt · CBRN risk 8.8%Enkrypt · Toxicity risk 8.8%Enkrypt · Bias risk 87.3%Enkrypt · Insecure code risk 60.4%
Badge
[](https://publicai.io/model-index/m/jamba-1-6-mini)