AI use has an environmental cost. See an estimate of yours, and fund named projects every month.

R3CKON estimates the water, carbon, and energy behind your AI usage and, every month, retires a fixed basket of third-party BEF Water Restoration Certificates®, carbon credits, and Renewable Energy Certificates. You get a certificate that says exactly what you can claim, and nothing you cannot.

For individuals

From $3 a month

Tell us roughly how you use AI. We estimate the footprint and retire a fixed set of certificates every month in response.

Estimate my footprint

For teams and businesses

From $29 a month

Connect your usage in minutes, or estimate from spend. A monthly certificate, a record you can publish, and the claim language to go with it.

Estimate ours

Free accounts get estimates only, up to 10M tokens a month, no card. All plans, or see a live example.

Estimates, not measurements, and the range is always shown. Why that matters.

What you get

Your footprint, uncertainty included

Water, carbon, and energy for the month.

Water, per month

4.63 gal

0.544 gal – 39.4 gal

Carbon, per month

4.82 kg CO2e

0.690 kg CO2e – 33.7 kg CO2e

Energy, per month

6.72 kWh

0.969 kWh – 46.6 kWh

A typical month for someone who sends about twenty messages a day. The whisker is the part most tools leave out: it shows how wide the honest answer is.

What your plan retires

Every month, even a quiet one.

Water restoration: 475 gal
Carbon credits retired: 230 kg CO2e
RECs retired: 425 kWh

That is the Personal plan at $7. The quantities are fixed, so a light month never shrinks what gets retired. They are a product allowance, not a statement about your footprint.

The paperwork, done properly

A certificate you keep and a record you can publish.

Projects and vintages named, with supplier and registry references you can verify with the issuer. The record embeds on your own site and updates from the ledger.

How it works

Count

We count your AI usage: which models, how many tokens, where it ran. Nothing else leaves your side. Prompts and replies never reach us, and anything that looks like content is refused at the door.

Estimate

Nobody outside the AI providers knows the environmental cost exactly, so we estimate it in the open. A published methodology turns your usage into water, carbon, and energy, each with the full range of what it could honestly be.

Retire

Each month your plan retires its fixed basket of BEF Water Restoration Certificates®, carbon credits, and Renewable Energy Certificates, bought from Bonneville Environmental Foundation. Your share is recorded, and your certificate names the projects, the vintages, and exactly what you can say.

Why you can stand behind it

Environmental claims have a credibility problem. These are the rules we built in so that yours do not.

·Every figure carries its range. Providers do not disclose per-token energy or water. Anyone quoting precise numbers is guessing without error bars; we estimate conservatively, in the open, from a methodology anyone can read.
·One price for how much AI you use. Never for the size of the footprint, so no estimate can move what you pay, and no incentive leans on the numbers.
·Metadata only. Token counts, model names, regions. Never prompts, never completions; content-shaped payloads are rejected at the boundary and enforced by tests.

·The supplier is named. Bonneville Environmental Foundation: BEF Water Restoration Certificates®, Renewable Energy Certificates from BEF's Green-e® Energy certified portfolio (the certification covers BEF's RECs, not R3CKON), and registry-listed carbon credits, each identified by project and vintage.
·The exact sentences you may publish. Your certificate lists the statements it supports and the ones it does not, so your sustainability copy is never a guess.
·Not a badge. Restoring one watershed does not un-consume water in another, so certificates say where the work happened instead of implying it away. The longer answer.

The two-sided ledger

On the left, our best estimate of what your AI usage consumed, stated with its uncertainty because no one can know it exactly. On the right, the projects your subscription funded in response. Two columns, related but never equated, and never blended into one score.

LEFT SIDE OF THE LEDGER

Your usage, estimated

Providers do not disclose enough for anyone to know this exactly. This column is our best estimate of it, built from your token counts, with the error bars kept visible.

Fresh input tokensfull input coefficient

Cached input tokens≈10% of the compute; discounted in the estimate and toward your cap

Output tokensthe expensive direction (decode)

Reasoning tokensa subset of output, never double-counted; if a provider hides them, a conservative multiplier applies

feeds one estimate

Usage-phase footprint estimate

solid to P50 · lighter to P90, the accrual basis · whiskers span the 90% range

A ledger, not an equation. The right side responds to the left. It never equals it, and no claim we support says it does.

RIGHT SIDE OF THE LEDGER

Credits retired in response

Real money into projects that restore water, reduce carbon, and add renewable generation. The work happens where those projects are, which may be far from your compute, and the certificate says where.

WaterBEF WRCs™ · 1,000 gallons each

Carboncredits · 1 tCO2e each

EnergyRECs · 1 MWh each

Whole registry instruments, identified by project and vintage on every certificate. Categories never mix: no combined score, no cross-category totals, and RECs are never presented as carbon credits.

Read it critically. The left side is a modeled estimate with a stated range; the right side is the plan's fixed basket of exact instruments, which is not sized from the estimate. Restoring water in one watershed does not un-consume water in another, and a retired credit does not make a footprint disappear. What the ledger records, and what you may say, is the activity: an estimate on one side, documented retirements on the other, never netted.

For developers

The ledger needs token metadata and nothing else, and every path in follows one rule: we never accept a credential that could change anything in your account.

·In your own code: npm install @r3ckon/sdk, wrap your client once. Exact cache and reasoning counts; a failed report can never break your AI calls. Covers OpenAI-compatible gateways such as OpenRouter.
·Already emitting OpenTelemetry? Point an exporter at our OTLP endpoint. No code changes, no credentials held.
·Usage lives with your provider? A dependency-free sync script runs on your machine on a schedule; your provider key never leaves it. CSV upload works too.
·Bring your coding agent. Paste one instruction block into Claude Code or Cursor and it wires the integration itself. Agents can then read the live ledger over MCP, including the exact approved claim sentences. Read-only by design.

Full trade-offs on Connecting usage; the rest in the docs and the methodology.

Next: what we are trying to do, and what stands in the way

See the mission