MCP.so
Sign In

Faim Time Series Forecasting

@S-FM

Repository

About Faim Time Series Forecasting

An MCP server for zero-shot time-series forecasting powered by foundation models such as Chronos 2 and TiRex, with support for multivariate and probabilistic forecasts.

#time-series#forecasting

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "faim": {
      "command": "npx",
      "args": [
        "-y",
        "@faim-group/mcp"
      ],
      "env": {
        "FAIM_API_KEY": "your-api-key-here"
      }
    }
  }
}

Tools

2

List all available forecasting models and their capabilities. Returns information about Chronos2, TiRex, and other available models, including supported output types and features.

Perform time series forecasting using FAIM platform. Supports both point forecasting (single value) and probabilistic forecasting (confidence intervals). Can handle univariate and multivariate time series data. Currently supported models: Chronos2 (default, recommended for multivariate) and TiRex (fast, univariate only).

Overview

What is Faim Time Series Forecasting?

The Faim Time Series Forecasting server is a Model Context Protocol (MCP) server that integrates the FAIM time series forecasting SDK with any MCP-compatible AI assistant. It enables AI-powered zero-shot forecasting using two foundation models: Chronos2 and TiRex.

How to use Faim Time Series Forecasting?

Install Node.js 20+, npm 10+, and obtain a FAIM API key from faim.it.com. Configure your MCP client to run the server via npx -y @faim-group/mcp or the global faim-mcp command, setting the FAIM_API_KEY environment variable. The server provides two tools: list_models to retrieve available models and forecast to perform time series forecasting with point, quantile, or sample outputs.

Key features of Faim Time Series Forecasting

  • Exposes two foundation time-series models: Chronos2 and TiRex.
  • Offers two MCP tools: list_models and forecast.
  • Accepts 1D arrays for single series or 3D arrays for batch inputs.
  • Supports point, quantile, and sample forecast outputs.
  • Allows custom quantile levels for risk assessment.
  • Works with any MCP-compatible AI assistant.

Use cases of Faim Time Series Forecasting

  • Perform zero-shot time series forecasting for any univariate series.
  • Run batch forecasts on multiple sequences using 3D input arrays.
  • Generate confidence intervals via quantile predictions.
  • Obtain distribution samples for probabilistic analysis.
  • Integrate forecasting into an existing MCP-connected AI workflow.

FAQ from Faim Time Series Forecasting

What models are available?

The server exposes Chronos2 and TiRex for zero-shot forecasting.

What input formats are supported?

Input can be a 1D array (single time series) or a 3D array in batch/sequence/feature format.

How do I get a FAIM API key?

Register at https://faim.it.com/ to obtain your API key, which must be set as the FAIM_API_KEY environment variable.

What are the system requirements?

Node.js 20+, npm 10+, and a valid FAIM API key are required.

Can I use this server with any LLM?

Yes, the server implements the standard MCP protocol and works with any application that implements an MCP client.

Frequently asked questions

What models are available?

The server exposes Chronos2 and TiRex for zero-shot forecasting.

What input formats are supported?

Input can be a 1D array (single time series) or a 3D array in batch/sequence/feature format.

How do I get a FAIM API key?

Register at [https://faim.it.com/](https://faim.it.com/) to obtain your API key, which must be set as the `FAIM_API_KEY` environment variable.

What are the system requirements?

Node.js 20+, npm 10+, and a valid FAIM API key are required.

Can I use this server with any LLM?

Yes, the server implements the standard MCP protocol and works with any application that implements an MCP client.

Comments

Basic information

Category

Productivity

Transports

stdio

Author

S-FM

Submitted by

Andrei Chernov

More Productivity MCP servers

Prose Coach

prose.coach

Deterministic writing filter. 43 AI patterns marked, each quoting the line that triggered it and attaching the fix. Connects as a remote MCP server at one URL. The free tier runs 3 drafts a day at 12,000 characters, no a

Fdback.io

Fdback

Fdback.io is a user feedback platform (feedback board, roadmap, changelog) with a built-in remote MCP server - no local install, connect any MCP client with a URL and API key. Your AI agents can: - List, search, and rea

My AskAI

My AskAI

Use My AskAI to find product, setup, integration, API, feature, pricing, and troubleshooting information. Search the public knowledge base, read complete documentation and OpenAPI files, and estimate pricing from monthly

Sitespeak Claude Plugin

sitespeakai

Manage your SiteSpeakAI chatbot from Claude Code - a plugin and MCP server for optimizing chatbots in plain English.

FlowyTeam OKR MCP Server

flowy-team

FlowyTeam is a native OKR MCP server. Connect Claude (Desktop, web, mobile), ChatGPT, Claude Code or n8n and your agent can read every objective and key result in your workspace, post check-ins, move key-result progress,