Introduction
Quantinal means "Quantized Signal" — turning abstract, hard-to-measure events into precise numerical scores you can track and act on.
Quantinal turns your questions into numbers you can act on.
Instead of asking ChatGPT once and getting a text answer, Quantinal monitors events continuously and gives you measurable scores from 0 to 100 that tell you exactly where things stand.
You want to track Tesla stock's trading signals. Instead of asking "How's Tesla doing?" every day:
- Create a Tracker that monitors Tesla sentiment
- Quantinal runs it automatically (hourly, daily, etc.)
- You get a score of 72 and a detailed insight
Quant
Quants are Quantinal's platform currency — 1 Quant = $1 USD. Everything on the platform is priced in quants: running trackers, storing data, and more. Quants power a usage-based, pay-as-you-go model with no hidden fees.
Space
A Space is your project workspace — a container that holds your trackers and results in one place. Each space is pre-allocated with storage (GB) and throughput (runs/hr) capacity.
Tracker
A Tracker is an AI-powered monitoring job that lives inside a Space. You tell it what to watch and how to score it, and it produces scored results on demand or on a schedule. Each tracker has a built-in Tracker Vault for uploading private data (PDFs, CSVs, JSON) that the agent uses as context during inference. Multiple trackers can share the same vault for common data.
Agent
An Agent is the AI brain that powers your tracker. You compose its behavior by selecting skills — composable building blocks like Web Research, Stock Research, and more. Control reasoning depth with Basic (×1), Advanced (×6), Super (×25), or Zen (×55, async only) modes.
Skills
Skills are composable building blocks that give your agent specialized capabilities. Select up to three when creating a tracker — they're additive, so combining multiple skills gives your agent broader coverage.
Scoring
When a tracker runs, it produces a scored result — a factor-based assessment from 0 to 100 that you can act on. The engine starts from a baseline (prior run or 50 on first run), applies evidence-weighted factor adjustments, and returns a structured insight explaining the move.
Composing Trackers
Link your own trackers (up to 10) so each run can build on scores and context from other monitors in the same workspace—deeper intelligence without leaving your account.
Pricing
All costs are in quants (1 quant = $1 USD). There are two types: Space capacity (monthly storage + throughput) and Tracker runs (per-run cost based on skills and reasoning mode). Data uploads are generally free.