A new kind of intelligence test IQ+ now in beta

IQ+ Beta evaluates how you think across 8 cognitive domains — from abstract reasoning to temporal intelligence. No time pressure, no single right answer. Real assessments from real people.

IQ+ Beta

IQ+ Beta is an experimental cognitive assessment designed to look at how people think, not just what they know.

There's something I've noticed over time

I can sit in front of a question, understand it completely, and still struggle to answer it.

But because I'm trying to work out what the person asking it actually wants.

So instead of answering naturally, I translate it.
Shape it into something more "acceptable".

And every time I do that, something gets lost.

Over time, that starts to add up.

You begin to wonder whether the way you think is the issue, or the way it's being measured.

The question behind it

IQ+ Beta came from a simple question:

What if we stopped asking for the "right" answer, and just looked at how someone actually thinks?

What the prompts reveal

Some of the prompts are deliberately open.

Example:

Describe the last room you were in.

Some people will list the basics.
Others go into texture, layout, sound, small inconsistencies, things most people filter out.

Those differences are worth noticing.

What IQ+ Beta does

IQ+ Beta turns that idea into a structured assessment designed to surface different kinds of thinking.

It allows you to:

See how your thinking shows up under different prompts
Understand which cognitive domains you naturally lean into
Recognise patterns in how you interpret, connect, and express ideas
Identify where your thinking is strong, and where it creates friction
Get a reflection that feels specific, not generic

It's not about getting the answer right.

It's about seeing how you arrive at one.

How it works

Step 1: Respond in your own way

Each task is designed to shift you into a different kind of thinking.

Some ask for pattern recognition
Some ask for interpretation
Some ask for reflection or construction

There is no single correct response.

You're not being tested on knowledge.
You're being observed for how you approach the problem.

And there is no time limit. That's a deliberate design decision. Penalising slow thinkers — particularly neurodivergent ones — tells you more about the test than the person taking it.

Step 2: Your thinking is evaluated across 8 domains

Instead of marking answers as right or wrong, IQ+ Beta looks at how your response expresses different types of thinking.

Analytical domains — where there are more and less correct answers:

Abstract Reasoning — how you work through unfamiliar problems with limited information
Pattern Fluency — how you recognise and articulate structure
Verbal Reasoning — how you interpret language, meaning, and inference
Logical Deduction — how you reason from evidence to conclusions

Interpretive domains — where what matters is depth, originality, and self-awareness:

Emotive Reasoning — how you read subtext, feeling, and human context
Meta-Insight — how you reflect on your own thinking
Creative Divergence — how you generate and expand ideas
Temporal Intelligence — how you reason across time, sequence, and change

Each domain captures a different signal.

Responses are evaluated using a defined rubric per domain — not open-ended AI opinion. The system looks for things like structure and coherence, depth of interpretation, originality, and self-awareness. Guided by structured criteria, applied consistently.

Step 3: You get a reflection, not just a score

You receive:

A short interpretation of how you tend to think
Relative strengths across domains
Insight into how your responses differ across task types

The overall score is your Cognitive Signature Index — a number out of 160. But the number matters less than the level it lands in. Those levels have names: Emerging, Engaged, Accomplished, Distinctive, Exceptional, Transcendent. They're chosen deliberately. No clinical language. Nothing that makes a lower score feel like a verdict.

It assumes that two people can arrive at completely different answers and both be demonstrating high capability.

Some think in patterns or systems.
Some think in images or stories.
Some think in emotion or abstraction.

None of those should be invisible just because they don't fit a standard format.

For the curious: how it actually works

Under the surface, IQ+ Beta combines a few elements:

Structured task design

Each task is mapped to a specific cognitive domain
32 tasks in the bank; each assessment draws 16 at random, two per domain
This ensures the assessment explores different thinking modes, not just one

Defined evaluation criteria

Each domain has explicit scoring logic
GPT-4o applies those criteria to responses

Consistent, explainable scoring

Generates both scores and qualitative feedback
Keeps evaluation consistent and explainable
Every scoring call is logged, including the AI's reasoning

Pattern interpretation

The system looks at:

Relative strengths across domains
Differences in responses across task types

From this, it generates an interpretation of how you tend to think, not just how you performed.

Feedback loop

Users can indicate whether the interpretation feels accurate
This creates a dataset to refine both tasks and scoring

The system improves based on real responses, not assumptions.

What this is (right now)

IQ+ Beta is live, and people are using it. There are 17 completed assessments in the dataset so far — real responses from real people, not simulations.

It's still evolving. But that's the point.

A tool for people who don't feel reflected in traditional measures
A way to surface overlooked cognitive strengths
A dataset for understanding how different kinds of thinking show up at scale
A starting point for more formal research

It only matters if it holds up in real use.

Who it's for

People who have taken tests before and felt they didn't reflect them
People who think in ways that don't always fit structured formats
People curious about how they actually think, not just how they perform

Try IQ+ Beta

Try it.
Break it.
Tell us what feels right, and what doesn't.

Because this only works if it actually sees you properly.