Orakle Pricing Model

Generate the most accurate fair market value
and forecasted values, up to 90 days.

A calibrated price for any electronics product, from one sentence

Orakle is a gen-AI pricing model. Give it a product description — for any electronics item, new or pre-owned, in natural language or as structured fields, for the most accurate output — and it generates a calibrated price distribution and a 90-day forecast for North America, Australia, UAE, and APAC. Other regions will be added upon request. The median is your point estimate. The spread between the P10 and P90 quantiles tells you how confident the market is. Orakle 0.1 is the current model release.

For analysts

Replace ad-hoc scraping and manual comp pulls with a single API call that returns calibrated quantile prices and a 90-day forward view.

For developers

One TypeScript call submits the job, streams progress over SSE, and resolves the final result.

For AI agents

The MCP server exposes a typed get_pricing tool to any MCP client. Drop pricing into Claude, Cursor, or your own agent in one command.

How Orakle works

  1. Step 1

    Describe the product

    Send a free-form description — “iPhone 15 Pro 128GB Grade B unlocked” — or structured fields for the most accurate output, plus the target market.

  2. Step 2

    Watch the synthesis

    Orakle pulls the live market, normalizes the product profile, and fuses signals through the pricing model. Progress streams over Server-Sent Events.

  3. Step 3

    Read the distribution

    You get quantile prices, the market currency, and a scenario-weighted forecast. Use the median for a point estimate, the spread for confidence.

Frequently asked questions

What does Orakle do?
Orakle is a gen-AI pricing model. Describe a product — natural language or structured fields — and it returns a calibrated quantile price distribution (P10, P50, P90) and a 90-day forecast across North America, Australia, UAE, and APAC. Other regions will be added upon request.
How is Orakle different from price scraping?
Scrapers return raw listings. Orakle generates a calibrated quantile distribution of fair market value and a 90-day forecast. It takes into account live market signals, listings, geo-political factors, trends, our proprietary internal market signals, and more.
How do I get access?
Access is gated by an account. Onboarding is currently manual — request access via the early-access form on this page or contact sales, and we’ll provision an account for you.
Once I have access, how do I use Orakle?
From the dashboard for ad-hoc pricing, the TypeScript SDK for application code, the MCP server for AI clients, or the REST API.
Which markets and product categories are supported?
Electronics, new and pre-owned, across North America, Australia, UAE, and APAC. Other regions will be added upon request. Established products are best covered; the distribution widens gracefully when the signal is weak.
How is pricing billed?
Self-serve tiers with included tokens plus metered overage, and Enterprise plans for higher volume. See the pricing page.
Is Orakle suitable for AI agents?
Yes — the primary use case. The MCP server exposes a typed get_pricing tool any MCP client can call. Docs are also served as raw Markdown at /llms.txt and /llms-full.txt for agent ingestion.