Skip to content
MCP server

P2Predict

Local MCP server that benchmarks part prices from your own purchasing history

Pricing
Free
Platforms
CLI, API, self-hosted
Version
1.1.0
Listing updated
today

Similar MCP servers

About P2Predict

P2Predict is a free MCP server for sales and crm. It is available on CLI, API and self-hosted.

The README says P2Predict learns from a team's past purchase records which factors, such as supplier, material, size, spec and region, move the price of a part. It then returns a target price for a new or proposed part, along with a likely range and a breakdown of how much each spec and the supplier contribute to that figure.

It runs as a local MCP server that plugs into an AI assistant the team already uses, so procurement and engineering staff ask questions in plain language rather than opening a separate dashboard. Training and prediction happen on the user's own machine, and the README states that purchasing data is not uploaded.

The project is aimed at procurement and engineering teams in sectors such as automotive, semiconductors and electronics. It is explicitly parametric price prediction, not a should-cost model, and it widens its range or asks for a quote when a part is unlike anything in the history it has seen.

What it does

Parametric price prediction
Estimates what a similar part should cost by learning from the prices a team has actually paid.
Per-spec cost attribution
Splits each predicted price into contributions from individual specs and the supplier.
Calibrated likely range
Attaches a likely price range to each estimate and flags where the data is too thin to rely on.
Local data processing
Trains and predicts on the user's machine, with purchasing data kept local according to the README.
Agent-based interface
Works through a connected AI agent in plain language instead of a separate app or dashboard.

Good for

  • Checking whether a supplier quote is in line with past purchase prices
  • Pricing the cost impact of a tighter tolerance or spec change during design review
  • Comparing the expected cost of the same part across different suppliers
  • Triaging a long RFQ to find the lines that need a closer look

Pricing

Free

Free for internal company use; commercial license needed for third-party services or paid products.

P2Predict FAQ

Is P2Predict a should-cost tool?

No. The README says it benchmarks against what the market has charged for comparable parts and does not build costs up from material, labor and machine time.

Does my purchasing data leave my machine?

According to the README, no. Training and prediction run locally, and only what the connected AI agent normally sends to its own model leaves the machine.

Which AI agents can use it?

The README says any agent that speaks MCP, including Claude, GPT-based agents, or a locally running model.

What license applies?

The README describes a source-available PolyForm Noncommercial 1.0.0 license with an added grant for internal company use. Commercial licensing is needed to serve third-party clients or embed it in a paid product.

Is P2Predict free?

Yes. P2Predict is free to use.

What are the best alternatives to P2Predict?

The closest alternatives to P2Predict are Maildignity, Calven, CompanyData, MisarReach MCP Server and Salesforce.

Where this comes from

  • Description: written from the project's README and its MCP Registry entry.

How AIPoweredBot checks servers