Benchmarks

It is the secure agentic AI platform agencies can't do their work without.

Deliverable quality score
88.9%
best other 75.8%
Head to head
92%
of judgments won

Deliverable quality score by family

FamilyJustinian3.2ChatGPTAstra 6.2ClaudeFable 5.2Grok4.7Gemini4 Carbon
Research3 use casesA-87.1%B77.3%C66.9%Not scoredC-61.4%
Permits3 use casesB+84.8%C+68.9%B-76.2%C62.4%C-61.3%
Compliance2 use casesA-87.7%B-76%B-76.7%C+70.3%D+54.5%
Grants3 use casesA93.9%B79.3%B78.1%Not scoredD+54.6%
Data2 use casesA94.5%B80.5%B77.2%B79.5%D-45.5%
IT3 use casesA-89%C+71.8%B-73.7%Not scoredNot scored
Workforce4 use casesA-87.6%B78%B79%D47.8%C-61.1%

Each cell averages the family's use cases. A from 92, A- from 87, B+ from 82, B from 77, B- from 72, C from 62, D from 47, F below 42.

Widest gaps

Where Justinian finished furthest ahead

The deliverables with the biggest gap between Justinian and the next AI.

Compliance and audit

Inspection tracker
  • Justinian
    92.3%
  • ChatGPT
    60.3%
  • Claude
    60.3%
  • Grok
    56.3%
  • Gemini
    39.7%
See the full results

IT and software

Accessible app development
  • Justinian
    94.7%
  • ChatGPT
    62.7%
  • Claude
    71.3%
  • Grok
    70.7%
  • Gemini
    47%
See the full results

Grants, contracts and budgets

Budget dashboard
  • Justinian
    97%
  • ChatGPT
    73.3%
  • Claude
    75.3%
  • Grok
    74%
  • Gemini
    69.3%
See the full results

Research and rules

Bill tracker
  • Justinian
    90%
  • ChatGPT
    70.7%
  • Claude
    51.7%
  • Grok
    11.7%No deliverable
  • Gemini
    63.3%
See the full results

Data and maps

Agency data dashboard
  • Justinian
    94.3%
  • ChatGPT
    75.7%
  • Claude
    76.3%
  • Grok
    73.3%
  • Gemini
    70.7%
See the full results

Grants, contracts and budgets

Grant application package
  • Justinian
    94.3%
  • ChatGPT
    78.3%
  • Claude
    72.3%
  • Grok
    68%
  • Gemini
    36.3%
See the full results

Security

Assessed for government security programs

Justinian for Government runs in AWS GovCloud (US). Agency data stays inside its own boundary and is never used to train AI models.

Deliverables by sector

What agencies make with Justinian

Apps, dashboards, automations, decks, documents and forms, graded on the files and tools themselves.

Research and rules

Permits and licensing

Compliance and audit

Grants, contracts and budgets

Data and maps

IT and software

Workforce and operations

Research that holds up

Where Justinian pulls away on research

The government's own datasets, checked against the official record.

Method

How we test

  1. 01

    Real agency deliverables

    One task per deliverable use case: trackers, dashboards, permit systems, forms, decks, workbooks, letters and one-pagers. Each carries its own rules and data, with problems planted where a careful reviewer would catch them.

  2. 02

    Every system can make files

    ChatGPT Astra 6.2, Claude Fable 5.2, Grok 4.7 and Gemini 4 Carbon run on their makers' own services with web search, high reasoning effort and a code tool whose files are what the user receives. When their code crashed or sent no files, they got up to two follow-ups. Justinian ran as shipped, with none.

  3. 03

    Graded as delivered

    Every file is rendered as the requester would open it and every app is opened in a browser. Code or instructions the user would still have to run, deploy or schedule do not count. Bill numbers and rule sections are checked against the official record, and every regulatory call is checked against the statute.

  4. 04

    Three blind judges

    Judges from OpenAI, Anthropic and Google, none of them a competing model, score each deliverable as the user would see it and rank all five systems with names masked. Every system is graded with the same checklist and scoring guide.

From the agencies

Agency staff rate it above the AI they already have

83%

rate it better than AI tools they have used

78%

say research tasks now take less time

77%

are likely to recommend Justinian

62%

say Justinian made new tasks possible

Briefing

Test Justinian on your agency's own work.

Request a briefing →

This page’s interactive features did not load. Reloading usually fixes it. You can email sales@vulcan.ai for demos or support@vulcan.ai for help.