Plans and Usage

Two products, clear windows, automatic resets.

Dipole ML, Inc. builds one platform: the raven model family, DCode, and the API that serves them. Access comes through two products, with pay-as-you-go credits for the API. This page explains how usage is measured and what each plan includes.

#Our Products

raven subscription (Free, Pro, or Ultra): powers the raven app and, at a reduced rate, DCode.
DCode Coding Plan: a separate subscription with our largest allowance, drawn first by DCode when you hold both.

The DipoleML API stands apart: it bills prepaid credits, pay-as-you-go, independent of every subscription. Subscriptions never touch credits, and credits never touch plan allowances.

#How usage is measured

Allowances are tracked per model against two windows: a 6-hour window and a weekly window. When a window empties it replenishes automatically; there is nothing to restart. Your dashboard shows current usage as a percentage per model, so you always know where you stand.

#What each plan includes

Plan	Allowance
Free	The baseline allowance on raven Flash, plus a one-time welcome credit to try raven Max.
Pro	At least 12x the Free allowance on raven Flash, with a matching step up on raven Max.
Ultra	Our highest allowance: about 5x Pro (roughly 60x Free) on raven Flash, and the largest step up on raven Max.
DCode Coding Plan	Our largest allowance overall, on both models, with a combined pool across the two and headroom built for all-day coding.

#Using DCode on a raven plan

Without the Coding Plan, DCode draws your raven allowance at a 30% rate: one unit of DCode usage deducts three-tenths of a unit from your raven allowance, because coding sessions reuse context heavily. With the Coding Plan, DCode draws that plan first and your raven allowance is untouched.

#API requests

API keys do not consume plan allowances at all. They debit prepaid credits per request at the published prices (see API Credits), with a per-minute rate limit as the only usage window.