> ## Documentation Index
> Fetch the complete documentation index at: https://docs.seneca.professionalaiagents.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

> Submit one question to a panel of independent models and read a structured research report.

Seneca submits one question to a panel of independent AI models and returns a structured report: the finding, claims, positions, disagreements, minority insights, uncertainties, attribution, and methodology.

Panel models do not browse the web, call tools, or use extended reasoning. Agreement in the report is a finding about that panel. It is not a human poll, and it is not proof that a claim is factually correct.

Call Seneca over REST at `https://seneca-research.fly.dev/v1`, or over MCP at `https://mcp.seneca.professionalaiagents.com/mcp`. Both use the same API key, panels, credit ledger, and stored report. Research is asynchronous. Create a run, poll until it finishes, then read the stored report. Original model text is a separate read.

<CardGroup cols={2}>
  <Card title="Quickstart" icon="rocket" href="/quickstart">
    Create a key, launch a run over REST, and read the report.
  </Card>

  <Card title="MCP" icon="plug" href="/mcp-server">
    Connect a Streamable HTTP client and call the six Seneca tools.
  </Card>

  <Card title="Authentication" icon="key" href="/authentication">
    Send a Seneca API key as a bearer token.
  </Card>

  <Card title="Research depths" icon="layers" href="/research/depths">
    Choose Wide, Consensus, or Seneca. The server owns the models and the price.
  </Card>

  <Card title="API reference" icon="code" href="/api-reference/overview">
    Generated from the public OpenAPI specification.
  </Card>
</CardGroup>

## What a run returns

<Steps>
  <Step title="A research run">
    `POST /v1/research` accepts the question and reserves the panel price. A new run returns `202`. Poll `GET /v1/research/{run_id}` until the status is `completed`, `partial`, or `failed`.
  </Step>

  <Step title="A customer report">
    `GET /v1/research/{run_id}/result` returns the stored report. Use `detail=compact` when another model will read it. The report does not include original model text.
  </Step>

  <Step title="Individual model responses">
    List stored responses, then read one model. Download the full panel as JSONL when you need every response for your own storage and analysis.
  </Step>
</Steps>

There are no webhooks in this version. Custom model selection is not part of this API. The operational integrations are the REST API and [MCP](/mcp-server).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.