API

Prompt-native developer access for hypotenuse inference.

Hypotenuse AI exposes geometry inference as a modern developer primitive with structured requests, response metadata, and integration patterns designed for product teams shipping at scale.

/v1/infer primary endpoint for geometry inference requests
JSON clean contracts for product clients and backend services
SDK-ready designed for browser apps, services, and partner integrations
Endpoints

A compact surface area built for product teams.

POST /v1/infer

Run a hypotenuse prediction

Send the highlighted side values, prompt context, and formatting expectations. Receive a structured prediction response.

POST /v1/prompts/preview

Preview generated prompt packages

Inspect the exact long-form request before routing it to a live model.

GET /v1/health

Check service readiness

Lightweight operational endpoint for status pages, CI checks, and deployment verification.

Request Shape

Every request carries inputs, precedent, and response rules.

Sample request payload

Clients send the side inputs, contextual examples, and output expectations that shape inference behavior.

{
  "input": {
    "side_a": 11,
    "side_b": 60
  },
  "examples": [
    {"a": 3, "b": 4, "answer": 5},
    {"a": 5, "b": 12, "answer": 13},
    {"a": 8, "b": 15, "answer": 17}
  ],
  "instructions": [
    "Infer the most plausible hypotenuse value",
    "Return confidence and a brief note",
    "Format the response as JSON"
  ]
}

Sample browser client

A product client can call the inference layer with a request like this.

const response = await fetch("/v1/infer", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify(payload)
});

const result = await response.json();
console.log(result.hypotenuse);
Response

What the model returns

  • `hypotenuse`: formatted numeric answer.
  • `confidence`: string or numeric confidence score.
  • `latency_ms`: timing for UI instrumentation.
  • `note`: product-facing explanation for downstream UX.
Integration

Where teams plug it in

  • Learning and tutoring platforms.
  • Spatial computing and technical design software.
  • Simulation, industrial, and robotics workflows.
  • Partner products embedding geometry into larger systems.
Reliability

Designed for teams that need geometry to behave like infrastructure.

Latency

Fast UI feedback loops

Response behavior is tuned for interactive products where users expect immediate spatial feedback.

Schema

Predictable payloads

Clean contracts reduce glue code and make the API easier to instrument across product and backend environments.

Governance

Production-minded controls

The platform is shaped for versioned contracts, monitored behavior, and operational visibility as usage expands.

Access

Need sandbox access, a pilot, or a deeper technical review.

The team works directly with partners integrating Hypotenuse AI into product, education, and spatial workflows.