Independent AI research

Intelligence,
before words.

We build models for the moment
understanding becomes a decision.

Introducing Hunch
Scroll to discoverPERCEPTION → JUDGMENT → ACTION
01OUR RESEARCH

A thought doesn’t
have to be
a sentence.

Language is one expression of intelligence. Judgment is another.

We study models that can read a situation, weigh the possibilities, and return a useful signal. Compact enough to work inside software. Explicit enough to be measured.

Our first model
01 / EFFICIENCY

Do the useful work.

Match the model to the decision. Spend computation where it changes the result.

02 / UNCERTAINTY

Keep the possibilities.

Return a distribution. Let the application choose what happens next.

03 / OPEN RESEARCH

Make it examinable.

Work toward releases with weights, evaluation methods, and documented limits.

02FIRST MODELRESEARCH PREVIEW
hunch

A model for
the moment
of decision.

Give Hunch a state, a question, and possible answers. Get a probability for every answer in one forward pass.

Explore the interface
A luminous red glass seed suspended above a dark stone, an abstract study of emerging intelligence.
HUNCH / STUDY 002AN IDEA TAKES SHAPE

State in.
Probabilities out.

Three primitives.
One programmable interface.

STATE

“I’ve been charged twice for my last order. Can you help?”

QUESTION

Where should this ticket go?

CANDIDATE PROBABILITIES
Billing0.91
Account0.06
Technical support0.03
ILLUSTRATIVE OUTPUT

A walkthrough of the interface, not live inference or an evaluation result.

1forward pass
0generated answer tokens
3question types
03RESEARCH NOTE

Judgment you
can inspect.

Hunch is in research preview. Its interface is simple; understanding where it works takes careful evaluation.

Read the model brief
01

The application stays in control.

Choose candidates, set thresholds, and decide when to ask for human review.

02

A probability is a claim to test.

Calibration can change across tasks. Evaluate the model on the situations where you will use it.

03

The limits are part of the work.

The preview explores routing and triage. Release evaluation and work on unfamiliar task families are ongoing.

04THE LAB

Farther out.
Closer to understanding.

Antares Labs is an independent research lab exploring efficient, useful machine intelligence.

We begin with judgment: how a model represents possibilities, how it expresses uncertainty, and how those signals become part of real software.

Hunch is our first model.

A NEW POINT OF VIEW.BACK TO THE TOP