# Seneca

- [Introduction](https://docs.seneca.professionalaiagents.com/index.md): Submit one question to a panel of independent models and read a structured research report.
- [Quickstart](https://docs.seneca.professionalaiagents.com/quickstart.md): Check the account, create a research run over REST, poll it, and read the compact report.
- [Authentication](https://docs.seneca.professionalaiagents.com/authentication.md): Create a Seneca API key and send it as a bearer token.
- [MCP](https://docs.seneca.professionalaiagents.com/mcp-server.md): Connect an MCP client over Streamable HTTP and call the six Seneca tools with a bearer API key.
- [Research depths](https://docs.seneca.professionalaiagents.com/research/depths.md): Wide, Consensus, and Seneca. The server chooses the models and the price.
- [Create a run](https://docs.seneca.professionalaiagents.com/research/create.md): POST /v1/research accepts a question, requires an idempotency key, and returns before the panel finishes.
- [Polling](https://docs.seneca.professionalaiagents.com/research/polling.md): Poll GET /v1/research/{run_id} until the run is completed, partial, or failed.
- [Results](https://docs.seneca.professionalaiagents.com/research/results.md): Read the stored customer report with detail=full or detail=compact.
- [Model responses](https://docs.seneca.professionalaiagents.com/research/model-responses.md): List stored model metadata, then read one model's text in a character window.
- [Full-panel export](https://docs.seneca.professionalaiagents.com/research/export.md): Download every stored model response as UTF-8 JSONL.
- [Billing](https://docs.seneca.professionalaiagents.com/billing.md): Credits are reserved when a run is accepted and charged when a customer report is delivered.
- [Errors](https://docs.seneca.professionalaiagents.com/errors.md): Public errors use error.code. Read that field before you retry.
- [Agents](https://docs.seneca.professionalaiagents.com/agents.md): Call Seneca over REST from a backend, or over MCP from an agent client.
- [API reference](https://docs.seneca.professionalaiagents.com/api-reference/overview.md): The public Seneca API, generated from the OpenAPI specification.
- [Create research](https://docs.seneca.professionalaiagents.com/api-reference/public-api/create-research.md): Accept a research question. A new run returns 202 and reserves the panel price. The same Idempotency-Key with the same prompt and panel returns 200 for 24 hours, sets Idempotent-Replayed: true, and does not reserve credits again.
- [Read research status](https://docs.seneca.professionalaiagents.com/api-reference/public-api/read-research-status.md): Poll a run after the create request has closed. In-progress statuses are queued, running, and synthesizing. Terminal statuses are completed, partial, and failed.
- [Read the research report](https://docs.seneca.professionalaiagents.com/api-reference/public-api/read-the-research-report.md): Return the stored customer report. detail=full, or omitting detail, returns that document. detail=compact removes the question object and places methodology last. Neither read starts research or writes a credit ledger row.
- [List stored model responses](https://docs.seneca.professionalaiagents.com/api-reference/public-api/list-stored-model-responses.md): Metadata for the model responses stored on a completed or partial run. Pages are ordered by the stored model id. The original text is not included. Pass next_cursor back as cursor. This read does not start research, call a provider, or charge credits.
- [Read one stored model response](https://docs.seneca.professionalaiagents.com/api-reference/public-api/read-one-stored-model-response.md): Original text for one model id stored on a completed or partial run. model_id is the id from the response list or from a report field such as analysis.minority_insights[].supporting_model_ids. Slashes in the id belong in this query parameter. The id is not rewritten when the model is later retired.…
- [Export stored model responses](https://docs.seneca.professionalaiagents.com/api-reference/public-api/export-stored-model-responses.md): Download every model response stored on a completed or partial run as UTF-8 JSONL. Each line is one model. text is the complete stored response, or null. Save the file for your own analysis. Do not load the whole panel into an AI agent's context. To inspect one finding, use /responses and /responses…
- [Read account](https://docs.seneca.professionalaiagents.com/api-reference/public-api/read-account.md): Plan, subscription status, available credits, entitlement, and monthly grant. A key can read this without an active subscription. Entitlement is required to launch paid research.
- [List research usage](https://docs.seneca.professionalaiagents.com/api-reference/public-api/list-research-usage.md): Workspace runs, newest first. Pass next_cursor back as cursor to read the following page. Each item includes credits_reserved, not the final charge.

## OpenAPI Specs

- [openapi](https://seneca-research.fly.dev/v1/openapi.json)


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