# Numerai

> Numerai is a hedge fund that runs a crowdsourced data-science tournament: participants download
> obfuscated financial datasets, train machine-learning models, and submit predictions each round,
> staking the NMR cryptocurrency on their models to earn payouts. Programmatic access is a single
> GraphQL API at api-tournament.numer.ai covering Numerai Classic, Numerai Signals, and Crypto Signals.

## API
- [GraphQL API](https://api-tournament.numer.ai): Single POST endpoint (Elixir/Absinthe), introspection enabled. Auth via API token pair (x-public-id / x-secret-key) with per-token scopes.
- [Documentation](https://docs.numer.ai): Tournament rules, data, submissions, staking, and automation guides.

## Specs & Artifacts
- [Authentication](https://raw.githubusercontent.com/api-evangelist/numerai/refs/heads/main/authentication/numerai-authentication.yml)
- [OAuth/Token Scopes](https://raw.githubusercontent.com/api-evangelist/numerai/refs/heads/main/scopes/numerai-scopes.yml): 8 token scopes (upload_submission, download_submission, stake, web3_staking, read/write user info, delete).
- [Rate Limits](https://raw.githubusercontent.com/api-evangelist/numerai/refs/heads/main/rate-limits/numerai-rate-limits.yml): 150 published per-operation limits (graphqlRateLimits).
- [Conventions](https://raw.githubusercontent.com/api-evangelist/numerai/refs/heads/main/conventions/numerai-conventions.yml)
- [Webhooks](https://raw.githubusercontent.com/api-evangelist/numerai/refs/heads/main/asyncapi/numerai-webhooks.yml): Compute/Prediction-Node round triggers.
- [Data Model](https://raw.githubusercontent.com/api-evangelist/numerai/refs/heads/main/data-model/numerai-data-model.yml)

## SDKs & CLI
- [numerapi (Python)](https://pypi.org/project/numerapi/): Docs-recommended client + CLI for datasets and submissions.
- [numerai-cli (Python)](https://pypi.org/project/numerai-cli/): Official CLI to deploy cloud Prediction Nodes.
- [numerai-tools (Python)](https://pypi.org/project/numerai-tools/): Official scoring/interaction tools.

## Docs
- [Getting Started](https://docs.numer.ai/tournament/learn)
- [Documentation](https://docs.numer.ai)
- [Bounties / Security](https://docs.numer.ai/numerai-tournament/bounties)
- [Status](https://numerai.statuspage.io)
- [GitHub](https://github.com/numerai)
- [Forum](https://forum.numer.ai)