‹ PublicAI Index
The LLM benchmark aggregator.
Mercury 2.5
Inception
Strongest in Cybersecurity (#40 of 47), weakest in Agents (#239 of 298). Above par in 0 of 11 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Fable 5.1 and ahead of Mistral Small 3.1 24B Instruct and Llama 4 Scout 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.
Reasoning39.9−25.5#169/1951/5
Mathematics34.7−29.1#143/1521/2
Coding38.7−27.5#174/1791/6
Agentic coding38−28.9#170/1731/5
Professional37.8−24.6#186/1941/2
Cybersecurity32.6−31.3#40/471/1
Legal43.3−22.6#136/1731/1
Medical32.4−34.1#145/1491/1
Finance35−27.4#156/1621/1
Agents43.6−22.8#239/2982/6
Knowledge work39.8−30.9#155/2082/2
§ 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 Mercury 2.5, left for the other.
§ 3 · Sources
Where the numbers come from
3 publications, 11 figures. Every one links to the page it was read from.
GDPval-AA 482
AA-Briefcase 378
Vals · Legal Research Bench 4.33%Vals · LegalBench 83.06%Vals · Finance Agent 18.89%Vals · MedCode 31.33%Vals · MedScribe 55.09%Vals · CyberBench 45.48%Vals · Vibe Code Bench 3.68% (OpenHands)Vals · Code Migration 4.52%Vals · ProofBench 3%
Badge
[](https://publicai.io/model-index/m/mercury-2-5)