Introducing simpler, more transparent pricing at Reducto

Starting September 1, 2026, we’re making some changes to how pricing works at Reducto.

The goal is straightforward: make it easier to understand what you’re paying, forecast what you’ll spend as your usage grows, and ensure that improvements in our models can translate more directly into better rates for customers.

For existing customers, the important part upfront is that your current usage will cost the same or less under the new structure.

Why we’re changing our pricing structure

When we first launched Parse, Reducto was largely one product, and a credit-based pricing system made sense.

Since then, Parse has become significantly more capable, and we’ve introduced products like Extract and Split with their own cost structures. That created a few problems.

It became harder to look at your usage and immediately understand what it cost. Forecasting spend often meant converting between credits, product multipliers, and different endpoint charges. And because discounts were applied through a shared credit rate, it was harder for us to price each individual product based on its actual economics.

We want pricing to be much simpler than that.

Clearer product-specific pricing

Going forward, one Reducto credit will always equal $0.01, so credits and dollars have a consistent relationship.

We’ll also communicate pricing through clear product-level rates, making it easier to see how much each Reducto endpoint costs and forecast spend based on your expected page volumes.

For endpoints like Extract and Split, we’re simplifying things further by moving to a single all-in per-page rate. Instead of calculating the endpoint cost and then adding the cost of Parse beforehand, you’ll have one rate that represents the full cost of processing the document.

Higher-volume customers will continue to receive discounted rates that scale with their usage.

Making model improvements easier to pass on

There’s another benefit to moving toward product-level pricing: we can price each endpoint based on its own underlying economics.

As our models become more accurate and efficient, we want those improvements to translate directly into better performance per dollar for our customers.

Extract is a good example. Previously, processing a standard document with Extract meant paying $0.03 per page for extraction plus $0.015 per page for parsing, for an all-in cost of $0.045 per page. With our upcoming Extract model, the price will be $0.02 per page all-in, with no additional Parse charge.

So in addition to making the price easier to understand and forecast, the new model will deliver better accuracy at less than half the previous all-in cost.

We expect to continue applying the same principle across Reducto: as our models improve, the economics of using them should improve too.

Easier to understand, easier to forecast

Ultimately, this change is about making Reducto pricing better reflect how customers actually use the platform.

You should be able to look at a product and understand what it costs. You should be able to forecast what your bill will look like at 100,000 pages or 10 million pages. And when we find ways to make our models more efficient, we should have a straightforward way to pass those improvements along.

We think this new structure gets us much closer to that.

You can find the complete updated pricing, API changes, and answers to common questions in our pricing documentation.