Skip to main content

Synthetic personas.
Real demographics.

77 statistically-grounded dimensions. Sub-20ms generation. Hierarchical dependencies that model real-world correlations. Feed realistic variety to any LLM workflow. At scale.

See the API →
John Donaldson
persona_7k2m9f4q1x8d3b6n
31 yrs
Male
Kentucky · US
Office Admin
Moderate
Education
Some College
Openness
Average
Income
$50k - $75k
Tech comfort
Comfortable
Height
5'11"

Married, two kids. Commutes by car to a small office in Lexington. Watches basketball most weekends.

Interests
BasketballTriviaAnimeTravel...6 more
+ 63 more dimensions in JSON
Generated
David Kowalczyk28 · Software Engineer · California, US
Marcus Eoin52 · Sound Designer · Edinburgh, UK
Aisha Rahman33 · Product Manager · London, UK
Chris Wood34 · Professional Athlete · Nottingham, UK
Callum Fraser41 · Operations Lead · Glasgow, UK
Ana Reyes22 · Student · California, US
Nia Evans27 · UX Designer · Cardiff, UK
James Kowalski55 · Factory Supervisor · Ohio, US
Siobhan Murphy52 · Nurse Practitioner · Belfast, UK
Tim Hecker50 · Sound Artist · Los Angeles, US
Dorothy Walsh58 · Nurse Practitioner · Pennsylvania, US
Harpreet Singh36 · Data Analyst · Manchester, UK
Luis Mendoza38 · Truck Driver · Texas, US
Ellie Morgan29 · Account Executive · Bristol, UK
David Kowalczyk28 · Software Engineer · California, US
Marcus Eoin52 · Sound Designer · Edinburgh, UK
Aisha Rahman33 · Product Manager · London, UK
Chris Wood34 · Professional Athlete · Nottingham, UK
Callum Fraser41 · Operations Lead · Glasgow, UK
Ana Reyes22 · Student · California, US
Nia Evans27 · UX Designer · Cardiff, UK
James Kowalski55 · Factory Supervisor · Ohio, US
Siobhan Murphy52 · Nurse Practitioner · Belfast, UK
Tim Hecker50 · Sound Artist · Los Angeles, US
Dorothy Walsh58 · Nurse Practitioner · Pennsylvania, US
Harpreet Singh36 · Data Analyst · Manchester, UK
Luis Mendoza38 · Truck Driver · Texas, US
Ellie Morgan29 · Account Executive · Bristol, UK
77
Demographic dimensions per persona
120k+
Statistical data points
<20ms
Typical server-side generation
99.9%
Uptime SLA
Why it exists

Random users
aren't random enough.

Real populations are messy. Income correlates with education. Geography shapes career access. Age influences tech habits. PersonaGen captures that complexity — 77 dimensions of statistically-grounded variety.

Ground AI in real demographics

Demographic context on demand. 77 statistically-grounded dimensions — from income to personality to physical appearance.

Realistic, not random

Surgeons don't earn minimum wage. Geography shapes opportunity. PersonaGen models real-world dependencies — not random attribute soup.

Population-scale in seconds

Generate 1,000 demographically-varied personas in seconds. Feed them to your LLM workflow. That's research scale without the research budget.

How it's built

Built from the
ground up.

Every persona is constructed through a five-stage dependency pipeline — earlier traits constrain later ones, then physical attributes are enriched into concrete measurements.

1
Foundation
Fixed at birth

Immutable traits that constrain everything downstream. Set first, never overridden.

Country of birthEthnicityGender
2
Life path
Shaped by Tier 1

Constrained by context of birth — parental background, regional opportunity, access.

Parental educationChildhood SESRegionImmigration
3
Career & income
Influenced by 1 + 2

Occupation and income constrained by education, location, and socioeconomic origin.

OccupationIncomeEmployment
4
Psychology & lifestyle
Shaped by all above

Personality, cognitive style, habits, and preferences for messaging and UX simulation.

Big FiveLiteracy & numeracyInterestsCommunication styleRisk tolerance
5
Physical profile
Derived + enriched last

Physical attributes generated after health and activity context, then enriched into measurements and sizing.

Body buildHair & eyesSkin toneHeight & weightClothing sizes
Try it

Generate a persona.
One in 10⁷⁹ quinvigintillion.

api.personagen.dev/v1/us/persona
Ready
Request
Country route
Unauthenticated limit (IP)
0 / 25
No API key: 25 requests per hour.
Sign in to create an API key for higher limits.
Press generate
77 dimensions · statistically grounded
sub-20ms server-side generation
Use cases

What you can
build with it.

Generate personas. Feed them to your LLM. Aggregate the output.

Research at scale
1,000-person focus group. No recruiting.

Generate 1,000 personas. Ask them all the same question. Aggregate responses, surface themes, cut by demographic.

Generating personas
0 / 1,000
For developers

One request.
Full persona.

One GET request returns a complete persona. US and UK routes available.

What's included
Dimensions77
CountriesUS, UK
Latency<20ms
Seed supportDeterministic
Demo limit25/hour per IP
API key5,000/month (more on request)
Request · bash
curl -X GET "https://api.personagen.dev/v1/us/persona" \
  -H "X-API-Key: pk_your_key_here"
Response · json (truncated)
{
  "success": true,
  "data": {
    "name":        { "// 2 fields" },
    "identity":    { "// 10 fields" },
    "origin":      { "// 7 fields" },
    "background":  { "// 3 fields" },
    "work":        { "// 7 fields" },
    "household":   { "// 12 fields" },
    "behavior":    { "// 7 fields" },
    "psychology":  { "// 9 fields" },
    "decisioning": { "// 5 fields" },
    "health":      { "// 2 fields" },
    "appearance":  { "// 10 fields + nested" }
  },
  "metadata": { "country": "us", "seed": "...", "version": "1.1.0" }
}
FAQ

Questions
worth asking.