TTSFM.ai
Public Docs
OpenAPI Source of Truth
MCP Streamable HTTP
CLI for Consumers

TSFM.ai developer documentation.

Multiple pages, one contract. API, MCP, and CLI are aligned on the same schema so teams can move from manual calls to production automation with zero drift.

Documentation

Everything you need to get started

New to time series foundation models? Start with the Learn section. Ready to build? Jump straight to the Quickstart or API Reference.

Learn

Understand time series foundation models before you integrate.

What are TSFMs?

Understand foundation models for time series — what they are, why they matter, and how they compare to classical methods.

Read more

How TSFMs work

Dive into architecture patterns, training objectives, and how time series data flows through transformers.

Read more

Use cases

Forecasting, anomaly detection, classification, and imputation across retail, energy, finance, and more.

Read more

Choosing a model

Decision framework and comparison table to select the right model for your latency, cost, and accuracy needs.

Read more

Benchmarks

How FEV Bench and GIFT-Eval rankings are sourced, parsed, and linked to hosted models.

Read more

Build

API reference, request schemas, and integration guides.

Quickstart

5 min

Register, validate, and make your first forecast call in under 5 minutes.

Open

API Reference

All HTTP endpoints — auth, models, forecasting, anomaly detection, and billing.

Open

Forecast Contract

Canonical request and response schema for the /v1/forecast endpoint.

Open

Four integration surfaces, one schema

The OpenAPI schema is the source of truth. Playground, API, MCP, and CLI are all aligned on the same contract.

Surface

Playground

Browser-based sandbox for running forecasts, anomaly detection, and more — no code required.

Best for exploration, prototyping, and quick data checks.

Open docs

Surface

HTTP API

Direct REST access for model discovery, auth, account controls, and forecasting inference.

Best for product backends, web apps, and native clients.

Open docs

Surface

MCP

Stateless Streamable HTTP server that exposes one tool per OpenAPI operationId.

Best for AI agents and automation tools.

Open docs

Surface

CLI

Schema-grounded CLI for discovery, inspection, and execution.

Best for operators, CI pipelines, and developer workflows.

Open docs

Typical request lifecycle

  1. 1Discover models via /api/models or the model catalog UI.
  2. 2Authenticate with session token or API key in Bearer format.
  3. 3Submit requests to /v1/forecast and related inference endpoints.
  4. 4Track usage and billing through /api/account/usage.
  5. 5Scale with MCP tools and CLI automation — same schema, no drift.

Quick links