All four layers
Watermark, metadata, classifier and the combined report. The only paid upgrade is the second classifier on deep assays, and it is opt-in per request.
Pricing
One credit, priced the same everywhere
There is a single unit of cost. A standard assay is one credit per started 1,000 words, a deep assay is ten, and a file adds one. Whether the request came from this page, the API or an agent over MCP makes no difference to the price.
Updated
The unit
What a credit buys
| Operation | Cost | Notes |
|---|---|---|
| Standard assay | 1 credit | Per started 1,000 words. Base classifier, watermark and metadata layers. |
| Deep assay | 10 credits | Per started 1,000 words. Adds the deep classifier and an agreement score. |
| File upload | +1 credit | Added once per file, for text extraction and the C2PA or XMP read. |
“Per started 1,000 words” means a 1,200-word article costs two credits on a standard assay, not 1.2. Rounding up is the honest way to describe it, so it is described that way here rather than discovered at the invoice.
Free
Free, without a card
The checker on this site is genuinely free. The limits exist because each assay costs us money at the classifier providers.
| Free tier | Daily allowance | Rate limit |
|---|---|---|
| Anonymous | 10 standard assays per day | 10 req/min |
| Signed in | 30 standard assays per day | 10 req/min |
Free assays are standard assays. They run the watermark, metadata and base classifier layers — the same layers a paid standard assay runs — and they do not touch a credit balance. Signing in triples the daily allowance and gives you history, exports and the API.
Packs
Credit packs, bought once
| Pack | Price | Per credit | What that buys |
|---|---|---|---|
| 500 credits | €19 | €0.038 | 500 standard assays of 1,000 words or fewer |
| 2,500 credits | €79 | €0.032 | 2,500 standard assays of 1,000 words or fewer |
| 10,000 credits | €249 | €0.025 | 10,000 standard assays of 1,000 words or fewer |
Packs suit uneven work: a newsroom that checks forty pieces one week and none the next, or an agency that runs a batch at the end of a campaign. No subscription, no monthly minimum, no clock.
Plans
API plans, for steady volume
| Plan | Price | Included monthly | Overage per credit | Rate limit |
|---|---|---|---|---|
| Developer | €29 / month | 1,000 credits | €0.04 | 60 req/min |
| Growth | €149 / month | 8,000 credits | €0.04 | 300 req/min |
| Scale | €499 / month | 40,000 credits | €0.03 | 1,000 req/min |
A plan includes a monthly credit allowance, a higher rate limit and overage that bills per credit rather than cutting you off. Plans are the right shape when assays are part of a pipeline that runs every day — content ingestion, moderation queues, an agent that checks its own output before publishing.
Every plan includes the full API surface and the MCP server from the first tier. We do not gate the API behind an enterprise conversation.
Included everywhere
What every tier gets
Exports with a hash
JSON and PDF, both carrying the SHA-256 of the input, the pipeline version and a UTC timestamp. Available from the free signed-in tier upward.
EU hosting and no retention
Frankfurt, and inputs are not stored by default on any tier — including the paid ones. Privacy is not a feature we sell back to you.
Questions
Pricing questions
Do credits expire?
Pack credits sit on the workspace balance and are drawn down as you use them; every movement is visible in the workspace ledger. Credits included in a monthly plan reset at the start of each billing period, so a plan is a commitment to a monthly volume rather than a stockpile.
What happens when I run out mid-pipeline?
The request returns insufficient_credits with a 402 and nothing is charged for it. Nothing is silently downgraded to a cheaper assay, and no assay is half-run and billed.
Why is a deep assay ten times the price?
It buys a second opinion from a different vendor with a different training set, and the deep classifier costs us more than ten times as much per word as the base one. The agreement score you get out of it is the most useful number in the report, and it only exists when two independent models have both looked.
Do you offer a free trial of the API?
Signing in grants a starter balance of credits so you can integrate against real responses before paying anything. There is no timed trial to expire.