API Keys & Security

Wevi API keys provide secure programmatic access to Wevi’s REST endpoints and Model Context Protocol (MCP) server.

Environments

Wevi provides two distinct execution environments:

1. Live (`wevi_live_...`)

Produces production-grade video renders. Quality follows your plan: free plans export 720p with a watermark, Pro and Enterprise export up to 4K without one.
Spends API & MCP credits (see below). Your plan's allowance is used first, then any purchased credits.
Used for customer-facing video generation, Zapier bots, and production apps.

2. Test (`wevi_test_...`)

Every scene render carries the Wevi watermark, regardless of plan.
Exports are capped at 720p (a higher quality is accepted but downgraded).
Renders and exports do not count against your monthly quota, but each key has a hard daily sandbox allowance (below).
Projects created with a test key stay sandboxed even when opened later in the web app; sandbox exports are labelled "(Sandbox)" and carry sandbox: true.
Ideal for local integration testing, CI/CD automated test pipelines, and MCP experimentation.

Sandbox daily allowance

Limit	Per test key	Resets
Scene renders	20 / day	00:00 UTC
Video exports	5 / day	00:00 UTC
Scenes per project	4	—

Unchanged scenes are reused between publishes and do not count again. When a cap is hit the API returns 429 with limit, used, requested and resetsAt, and the MCP tool relays that message. Switch to a wevi_live_ key for production volume.

API & MCP credits

Every action a live key performs is metered in credits. Web sessions and sandbox keys never spend credits.

What costs what

Action	Credits
Scene render (new or changed scene)	1
Export at 720p	1
Export at 1080p	2
Export at 4K	4
Storyboard draft (`POST /ai/storyboard/draft`)	1
Web UI capture (`POST /browse/capture`)	1

A 3-scene 1080p video costs 5 credits. Re-publishing a project re-renders only changed scenes, so retries are cheap.

Included in each plan

Plan	Included credits	Reset
Free	10	Lifetime (one-time allowance to try MCP)
Pro (monthly or yearly)	150 / month	With your billing period
Enterprise	Unlimited	—

Top-up packs

Buy more from Profile → API Keys & MCP → Buy credits (or the pricing page). Packs are one-time purchases, never reset, and are used after your plan allowance. Free users can buy a pack to use MCP without subscribing.

The packs on sale and their prices can change; the current list always comes back from GET /billing/credits (packs) and the wevi_get_credits tool. At the time of writing:

Pack	Credits	Price
Starter	50	$15
Builder	200	$49
Scale	1,000	$199

When credits run out

The API returns 402 Payment Required:

{
  "statusCode": 402,
  "error": "Insufficient credits",
  "message": "This action needs 3 credits but 1 is available. Buy a credit pack or upgrade your plan at https://app.wevi.ai/app/profile?id=api-keys&topup=1.",
  "creditsRequired": 3,
  "creditsAvailable": 1,
  "planRemaining": 1,
  "purchasedBalance": 0,
  "topUpUrl": "https://app.wevi.ai/app/profile?id=api-keys&topup=1"
}

The MCP wevi_get_credits tool and GET /billing/credits return your balance, per-action costs and the packs on sale, so an agent can check before starting a large batch.

Where usage shows up

Your dashboard (Profile → API Keys & MCP): plan credits used, purchased balance, spend per key, sandbox allowance for today, and recent API activity.

Key limits & lifecycle

Rule	Value
Active Live keys per account	10
Active Test keys per account	5
Revocation	Immediate; the next request gets `401`
Expiry	Optional date set at creation; expired keys are refused like revoked ones
Retention	Revoked and expired keys are kept on record with their usage history. The dashboard hides them after 30 days behind a "Show older revoked" toggle

Security Architecture

One-Time Secret Display: When an API key is generated, the plaintext secret is revealed only once. Wevi does not store the raw secret key.
SHA-256 Hash Storage: The backend hashes keys with SHA-256 before storing them in PostgreSQL.
Instant Revocation: Revoking a key from your dashboard immediately disconnects all active Claude Desktop instances and API clients.

Rate Limits

Environment	Rate Limit
Standard API	120 requests per minute per IP/key
Render Export Trigger	Concurrency controlled by plan tier (Free: 1 concurrent, Pro/Enterprise: high concurrency)

If you exceed 120 requests/minute, the server responds with HTTP status 429 Too Many Requests.

API Keys & Security

On this page