# Mantic docs

> Add .md to any page URL for its Markdown, e.g. /read-forecasts/latest.md.

- [Overview](/): Learn how Mantic forecasts work and build a reliable integration with the Mantic API.
- Get started
  - [Read your first forecast](/getting-started/quickstart): Find a forecast, read its latest run and retrieve the reasoning behind it with a short Python script.
  - [Get and refresh an access token](/getting-started/authentication): Exchange your client ID and client secret for an access token, refresh it before expiry, and retry once on 401.
  - [Page through results](/getting-started/pagination): Follow the after and next cursor contract without skipping results or looping forever.
  - [Handle errors and limits](/getting-started/errors-and-limits): Respond to Mantic API error status codes, retry safely and honour polling and rate-limit headers.
  - [Beta features](/getting-started/beta): Newer endpoints and fields, available under the /beta path.
  - [Use these docs with an AI assistant](/getting-started/ai-assistants): Give an AI assistant these docs as Markdown, an llms.txt index or an MCP server.
- Core concepts
  - [Understand a forecast](/concepts/reading-a-forecast): Interpret a forecast's probability, uncertainty, reasoning and resolution criteria together.
- Find forecasts
  - [Search forecasts by text](/find-questions/search): Find forecasts about a topic when you don't know their title or ID.
  - [Filter and sort forecasts](/find-questions/filter-sort): List forecasts with chosen tags, ordered by latest run or creation time.
  - [List forecasts on a dashboard](/find-questions/dashboards): List dashboards and read their forecasts, each with its latest run.
- Read forecasts
  - [Read a forecast's latest run](/read-forecasts/latest): Read a forecast's latest run by type, and the fields that define how the forecast resolves.
  - [Get the probability above or below a value](/read-forecasts/probability-thresholds): Sample a forecast run's cumulative distribution to get the probability at or below, or above, an exact numerical or date value.
  - [Build a forward curve from several forecasts](/read-forecasts/forward-curve): Line up numerical forecasts for successive dates into one curve, with its median and ranges at each date.
  - [Read reasoning and sources](/read-forecasts/reasoning-sources): Fetch one forecast run's Markdown reasoning, with its source links.
  - [Compare forecast runs over time](/read-forecasts/history-delta): Measure how a forecast moved between runs, and detect when a new run completes.
  - [Check past forecasts against outcomes](/read-forecasts/past-forecasts): Rebuild what a numerical forecast said on each day before its outcome, and measure how often the outcome fell inside its ranges.
- Economic releases
  - [Forecast the next release](/economic-releases/next-print): Find a country's upcoming economic releases, read a release's forecast, and compare it with consensus, the prior value and the published actual.
- Scenarios
  - [Compare conditional outcomes](/scenarios/conditional-outcomes): Compare a forecast's yes and no scenario branches, and sample both conditional distributions at the same value.
  - [Forecast with your own context](/scenarios/your-own-context): Run a forecast that takes your facts or assumptions into account, wait for it to complete, and compare it with the forecast without them.
- Create forecasts (beta)
  - [Draft, validate and create a forecast](/create-questions/draft-validate-create): Create a draft, replace it until it is ready, then create the forecast once.
  - [Create forecasts in bulk from a template](/create-questions/bulk-template): Validate a CSV of company names, preview every rendered draft, then create the forecasts one at a time with a resumable checkpoint.
- Example workflows
  - [Build a dashboard briefing](/assemble/dashboard-briefing): Build a Markdown brief one step at a time, from a dashboard's forecasts to their movement between runs.
  - [Compare Mantic with prediction markets](/assemble/divergence-screen): Pair Mantic event forecasts with Polymarket markets, compare Mantic's probability with the market midpoint, and rank the largest gaps.
- API reference
  - [API reference](/api-reference): Authenticate with Mantic and explore every endpoint, request, and response.
  - Dashboards
    - [Get dashboards](/api-reference/dashboards/get_dashboards_dashboards_get)
    - [Get dashboard questions](/api-reference/dashboards/get_dashboard_questions_dashboards__dashboard_id__questions_get)
  - Economics
    - [Get indicators](/api-reference/economics/get_indicators_economics_indicators_get)
    - [Get indicator](/api-reference/economics/get_indicator_economics_indicators__indicator_id__get)
    - [Get indicator releases](/api-reference/economics/get_indicator_releases_economics_indicators__indicator_id__releases_get)
    - [Get releases](/api-reference/economics/get_releases_economics_releases_get)
    - [Get release](/api-reference/economics/get_release_economics_releases__release_id__get)
    - [Get release forecasts](/api-reference/economics/get_release_forecasts_economics_releases__release_id__forecasts_get)
  - Forecasts
    - [Get forecasts](/api-reference/forecasts/get_forecasts_forecasts_get)
    - [Get forecast](/api-reference/forecasts/get_forecast_forecasts__forecast_id__get)
    - [Get forecast CDF](/api-reference/forecasts/get_forecast_cdf_forecasts__forecast_id__cdf_get)
  - Questions
    - [Get questions](/api-reference/questions/get_questions_questions_get)
    - [Get question](/api-reference/questions/get_question_questions__question_id__get)
    - [Get question forecasts](/api-reference/questions/get_question_forecasts_questions__question_id__forecasts_get)
    - [Get question scenarios](/api-reference/questions/get_question_scenarios_questions__question_id__scenarios_get)
  - Scenarios
    - [Get scenario cdfs](/api-reference/scenarios/get_scenario_cdfs_scenarios__scenario_id__cdfs_get)
  - Tags
    - [Get tags](/api-reference/tags/get_tags_tags_get)
  - Beta API reference
    - [Beta API reference](/api-reference/beta): Endpoints available only under the /beta path.
    - Dashboards
      - [Add dashboard question](/api-reference/beta/dashboards/add_dashboard_question_dashboards__dashboard_id__questions_post)
      - [Remove dashboard question](/api-reference/beta/dashboards/remove_dashboard_question_dashboards__dashboard_id__questions__question_id__delete)
    - Question drafts
      - [Get question drafts](/api-reference/beta/question-drafts/get_question_drafts_question_drafts_get)
      - [Create question draft](/api-reference/beta/question-drafts/create_question_draft_question_drafts_post)
      - [Get question draft](/api-reference/beta/question-drafts/get_question_draft_question_drafts__question_draft_id__get)
      - [Delete question draft](/api-reference/beta/question-drafts/delete_question_draft_question_drafts__question_draft_id__delete): Hide a draft from the caller for good, keeping its record and any question made from it.

Raises:
    HTTPException: 404 when the caller has no live draft with this id.
      - [Replace question draft](/api-reference/beta/question-drafts/replace_question_draft_question_drafts__question_draft_id__put)
    - Question sets
      - [Create question set](/api-reference/beta/question-sets/create_question_set_question_sets_post): Create a question from each draft and group them, in one transaction.
      - [Get question set](/api-reference/beta/question-sets/get_question_set_question_sets__question_set_id__get)
      - [Get question set questions](/api-reference/beta/question-sets/get_question_set_questions_question_sets__question_set_id__questions_get)
      - [Get question set runs](/api-reference/beta/question-sets/get_question_set_runs_question_sets__question_set_id__runs_get): List every Scout run of the set, with each completed run's forecast.
    - Questions
      - [Create question](/api-reference/beta/questions/create_question_questions_post)
      - [Create question forecast](/api-reference/beta/questions/create_question_forecast_questions__question_id__forecasts_post): Start a forecast of the question that takes `additional_context` into account.

Give facts with their source and date where you can, and label assumptions as
hypothetical: the forecast weighs stated facts as evidence, treats labelled assumptions
separately, and keeps the question and its resolution criteria unchanged.

The forecast builds on the question's latest forecast, is marked `user_context` and is
visible to you and your organisation. `GET /forecasts/{id}` returns it once it completes.
    - Search
      - [Search](/api-reference/beta/search/search_search_get): Search questions, dashboards and situations, grouped by kind; omit `q` to browse.