← Blog · Pricing repeat workflows · May 1, 2026 · Rhumb · 9 min read
Answer target: price agent workflow

Price the first repeat agent workflow before you wire it into a loop.

The dangerous moment is not the first demo call. It is the second, tenth, or hundredth call after the agent starts retrying, falling back, enriching, and recovering without a human watching the bill.

Current truth boundary

Rhumb's discovery surface covers 999 scored services and 435 capability definitions, but current callable execution is narrower at 18 callable providers strongest for research, extraction, generation, and narrow enrichment. Price the workflow that can actually run today, not the whole catalog.

Start with the repeat job, not the platform idea

"How much will agent execution cost?" is too broad to answer honestly. The useful unit is one repeated, observable job: search for ten fresh sources, extract one vendor page, send one approved email, generate one image, verify one lead, or classify one document.

Once that job is concrete, the price model becomes operational: resolve the supported capability, estimate the route, choose the credential rail, cap repeat volume, test the denied neighbor, and keep receipt evidence that survives retries.

The six fields to price before launch

Capability

What exact job will the agent repeat?

Use a concrete capability id such as search.query, email.send, image.generate, scrape.extract, or the current supported equivalent — not a broad product idea.

Provider path

Should Resolve choose the best supported provider, or should the agent pin one provider?

A pinned provider is a constraint, not a price shortcut. Estimate, credential mode, denial, and receipt checks still apply.

Repeat volume

How many times can the agent call before a human reviews the result?

Name per-run, per-hour, and per-day ceilings so the first loop cannot turn a good demo into a quiet spend spiral.

Cost ceiling

What is the maximum acceptable cost for one completed action?

The ceiling should include retries, fallbacks, enrichment calls, and output expansion — not just the happy-path provider price.

Credential rail

Whose account, key, wallet, or quota pool pays when the workflow repeats?

Governed key, wallet-prefund, x402, BYOK, Agent Vault, and provider pinning create different budget evidence.

Denied neighbor

What adjacent action must fail closed before the loop is trusted?

If you cannot name the denied domain, tenant, amount, row, account, path, or side effect, the price is not production-real yet.

Choose the rail after the workflow is priced

Governed Rhumb key

Best when

Default for repeat managed execution when the workflow fits Rhumb's current callable surface.

Pricing shape

Upstream execution cost plus 20% margin under Rhumb-managed billing.

Evidence to keep

Run resolve and execute estimate before the first paid call, then keep the estimate id, capability id, credential mode, and cost ceiling in trace context.

Get governed key
Wallet-prefund

Best when

Wallet-native agents that will call repeatedly but should not authorize every request separately.

Pricing shape

Wallet funds reusable balance, then repeat traffic executes on the stable X-Rhumb-Key rail.

Evidence to keep

Price the workflow like governed execution, but preserve wallet identity, top-up id, amount, and balance drawdown in the receipt trail.

Review wallet-prefund
x402 per-call

Best when

Zero-signup or strict per-request payment flows where payment proof is part of the product experience.

Pricing shape

USDC on Base; upstream execution cost plus 15% margin on the x402 rail.

Evidence to keep

Expect payment instructions per request and prove payment_request_id, X-Payment proof, amount, wallet identity, and verification outcome before execution state changes.

Review x402
BYOK or Agent Vault

Best when

Workflows that must run through your provider account, contract, workspace, or compliance boundary.

Pricing shape

Rhumb adds no markup in BYOK or Agent Vault mode. Provider charges bill directly to your provider account. You still pay the provider directly.

Evidence to keep

Estimate Rhumb-side routing and execution shape, then separately cap the provider account's quota, rate limits, and per-action cost exposure.

Choose credential path
Estimate before execution

The estimate call is the budget checkpoint

The pricing page exposes the canonical estimate shape at https://api.rhumb.dev/v1/capabilities/{capability_id}/execute/estimate. Use it before promotion, with the same capability id, provider constraint, credential mode, and budget ceiling the agent will use in production.

The estimate is not just a number. It is the checkpoint where route, availability, credential mode, cost ceiling, policy constraint, and denial expectations become explicit enough to compare against the eventual receipt.

Loop math: price the completed action

Happy-path provider cost is only the first term. Add retries, fallback providers, search/extraction enrichment, validation reads, and output-processing calls before you call it costed.
Use a cost ceiling per completed user-visible action, not only per raw provider call. One successful action may include several API calls.
Separate fail-free reads from billable execution. Discovery, scoring, and browsing are free; side-effecting or provider-routed execution is the paid surface.
Provider errors should not become billable successes. The receipt needs a typed outcome so finance, retry, and recovery logic do not guess.
If a retry resumes a completed-but-unacknowledged write, idempotency must preserve the original side effect instead of buying a second one.

Ready-to-loop checklist

A concrete capability id and supported provider path are known.
The estimate endpoint has been called for the exact credential mode and provider constraint you plan to use.
A human-readable cost ceiling exists for one completed action and for one unattended loop window.
The credential rail names the budget owner: Rhumb-managed account, wallet balance, x402 payment proof, provider account, or vault-bound connection.
The denied neighbor fails before provider execution, billing, or side effects observe it.
The receipt preserves capability id, provider, credential mode, estimate/cost fields, idempotency key, denial reason, and recovery state.

Anti-patterns

Pricing a broad project such as 'automate sales ops' before pricing one repeat job such as 'enrich one inbound lead from these two sources'.
Using a leaderboard score as a cost model. Scores help choose candidates; estimates and receipts price execution.
Letting an agent choose a cheaper-looking provider after the first failure without a fallback cost ceiling.
Treating BYOK as free because Rhumb adds no markup while ignoring the provider contract, quota, and retry spend.
Using x402 for repeat production traffic when the actual need is durable budget ownership and operator-visible controls.
E-006 proof sprint

Ask for proof around one priced workflow

The best request is narrow: one capability, one repeat volume, one credential rail, one denied neighbor, and one receipt shape you would trust after a retry. That is enough to test whether the first paid call can be boring.