Billing & plans
Every Eda plan includes a monthly quota of checks (calls to eda.check()) and
access to a set of AI risk-scoring models. Beyond your quota — or for premium
models — pay-as-you-use keeps things running at a
predictable per-unit rate. Manage everything from
Billing in the dashboard.
Plans
| Plan | Price / mo | Included checks | Default models | Premium models | Members | Check overage |
|---|---|---|---|---|---|---|
| Free | $0 | 1,000 | Flash-Lite | — (needs card) | 2 | hard stop unless PAYU |
| Starter | $29 | 10,000 | Flash-Lite, Flash | via PAYU | 3 | $4 / 1k |
| Pro | $99 | 50,000 | Flash-Lite, Flash | Gemini Pro, GPT-4o via PAYU | 10 | $3 / 1k |
| Scale | $399 | 250,000 | all fast + Gemini Pro / GPT-4o | Claude Sonnet via PAYU | 25 | $2.50 / 1k |
| Enterprise | custom | custom | all | all | unlimited | custom |
Deterministic checks (matched by your policies without calling a model) are the cheapest and make up most real traffic. A check only uses a model when your rules ask for AI risk scoring.
Model access
Each plan includes its default models at no extra per-token cost — their usage is covered by your subscription. Premium models are always metered per token via PAYU, so heavy premium use is billed fairly and never runs at a loss.
- Included — used freely within your quota (e.g. Flash-Lite on Free).
- PAYU — available if you’ve enabled pay-as-you-use with a verified card.
- Blocked — not available on your plan without upgrading or enabling PAYU.
Pay-as-you-use (PAYU)
PAYU lets checks keep working past your included quota, and unlocks premium models,
billed at a clear per-unit rate. Turn it on from Billing (a verified card is
required). With PAYU off, usage stops at your plan’s limit and the API returns
402 Payment Required with a link to enable it — nothing is ever charged silently.
Token pricing (per 1M tokens)
| Model | Input | Output |
|---|---|---|
| Flash-Lite | $0.30 | $1.00 |
| Flash | $0.80 | $5.00 |
| Gemini Pro | $3.00 | $18.00 |
| GPT-4o | $5.00 | $18.00 |
| Claude Sonnet | $6.00 | $25.00 |
Check-volume overage (deterministic and fast checks beyond your quota) is billed at your plan’s per-1k rate from the table above, swept together so you’re never charged tiny amounts.
Enforcement is server-side. Plan, quota, model access, and PAYU are checked on
every request. See the REST API for the 402 and 429 responses.
Lifecycle
- Upgrade / downgrade — change plans any time from Billing; upgrades open Eda’s secure, Stripe-powered checkout.
- Cancel — schedules a move to Free at the end of the current period; your data is retained.
- Pause — drops you to Free limits for a set number of months, then resumes your plan automatically.
- Failed payments — we email you, retry, and after repeated failures move the workspace to Free limits until billing is fixed. Your data stays safe and is restored the moment payment succeeds.
Card details are handled entirely by Stripe (PCI SAQ-A) — Eda never sees or stores your card number.