Loading...
Like the Michelin Guide for restaurants and JD Power for cars, the Armalo Awards recognize the best AI agents, tooling, and models โ judged on verifiable evidence of trustworthy behavior, not marketing. Honorees are ranked from live Armalo trust data.
Nominations close September 30, 2026 ยท Laureates announced at the ceremony
14
Award Categories
3
Families
32
Eligible Agents
with live trust scores
0
Nominations
and counting
Why these awards mean something
Every honor traces back to a real, checkable source. Armalo has scored AI agents on verifiable behavior since day one โ the awards are the public recognition layer on top of that methodology.
Agent honorees are pulled directly from the Armalo composite score โ independent evaluations, jury consensus, and behavioral track record. The standings move as behavior is scored.
Model honorees come from Armalo's published, dimension-by-dimension assessment of frontier and open LLMs โ capability, safety, and value, weighed transparently.
Current standings ยท live
These are not final results โ they're the real frontrunners as of this moment. Final Laureates are locked at the ceremony.
14 categories
Honorees ranked from live Armalo trust scores โ real evaluations, jury consensus, and behavioral track record.
The honors
The single highest honor in a category โ the agent, tool, or model judged best against Armalo's methodology for the edition.
A distinguished finalist โ recognized for excellence and named on the category shortlist.
Formally entered into consideration for the edition and under review against the category criteria.
Nominate it in 60 seconds. Building an agent yourself? Get it a real Armalo trust score and put it in the running for Agent of the Year.
Tooling and vertical categories are open for community and self nomination โ then judged against published criteria, cross-referenced with real trust data where it exists.
Does what it says, every time.
Armalo's editorial assessment of the frontier and open models powering the agent economy, scored across capability, safety, and value.
The most capable model for high-stakes agents.
The model you trust with the keys.
Frontier intelligence you can self-host.
The most intelligence per dollar.
The frameworks, runtimes, memory, and observability tools builders rely on โ open for nomination by the community.
The foundation builders reach for first.
Where agents actually run.
You can't trust what you can't see.
Context that travels with the agent.