Public integration reference

One research system.
Two protocols.

Use REST when your product controls the workflow. Use MCP when an AI client should discover and operate Minds as a tool. Both lead to the same durable research infrastructure.

DEFINE AUDIENCESGROUND GROUPSRUN PANELSRETRIEVE RESULTS

Protocol decision router

What are you building?

Pick the job. The lab returns the shortest supported route and a maintained source to continue from.

RECOMMENDED ROUTEREST API

Keep orchestration in your product.

Call typed public endpoints from your backend, keep API keys server-side, and store durable run identifiers with your application state.

Runtime source of truth

The contract is live.

This page does not maintain a second endpoint catalog. Counts and resource groups are calculated from the production OpenAPI document when you load it.

OPENAPI3.1.0contract version
PATHS24public routes
OPERATIONS29callable actions
RESOURCES7capability groups
LIVE RESOURCE GROUPSLoading production schema…
GroupsSparksPanelsResearchRuns

A durable research loop

The protocol changes.
The workflow does not.

  1. 01
    Define the audience

    Describe the market, customer, or stakeholder population the study must represent.

  2. 02
    Ground the groups

    Create explicit cohorts and reusable research context before asking the first question.

  3. 03
    Launch the study

    Run panels and research jobs as durable work rather than holding a chat session open.

  4. 04
    Retrieve the output

    Poll or resume by run identifier, then bring structured findings into the system making the decision.

Copy a starting point

Start at the transport layer.

REST / OPENAPI TYPES
npx openapi-typescript \
  https://api.getminds.ai/api/v1/openapi.json \
  --output minds-api.d.ts

Minds developer infrastructure

Move from a research question
to a durable result.