BUSINESS PLAN

Business - $49 per agent per month.

For agents at production volume. Audit logs, advanced policies, team access, $100k Spend Permission cap, and the lowest outflow fee we offer. Save 2 months on annual billing.

$49per agent / month

or $490/year (~$41/mo equivalent, 2 months free)

0% inflow fee, 2% outflow fee

Business is for agents that move real money or matter to a real team. The 2% outflow fee makes economic sense when monthly outflow crosses roughly $1,000 per agent (~$850 on annual billing). Below that, Pro is the better fit.

WHAT IS INCLUDED

Everything on Pro, plus the controls operators ask for.

Everything on Pro
2% outflow fee (lowest available)
Spend Permission cap up to $100,000 per agent
Full audit log of every policy decision, every API call, every webhook delivery
Recipient allowlist mode (restrict agent payments to specific addresses or verified merchants)
Custom Open Graph images per agent
Higher API rate limits (10x Pro)
Agent-level seat sharing (3 team members can manage one agent)
Priority email support, typical response within 4 hours
Slack support channel at scale (negotiated separately)

WHO BUSINESS IS FOR

Built for teams running money-moving agents.

Web3 automation agencies managing many client agents
Mid-market companies running AI agents in customer-facing roles
Operators with compliance or audit requirements
Anyone processing $1k+/month per agent in outflow volume (the Pro -> Business break-even)

WORKSPACE ADD-ONS THAT PAIR WITH BUSINESS

Scale without leaving Blockchain0x.

Extra seats at $5/seat/month for larger teams
Multiple Business agents in one workspace for parallel client work

BUSINESS VS PRO - THE MATH

When the 2% outflow fee pays for the $20 subscription difference.

The strict crossover from Pro to Business is $1,000/month per agent in outflow volume. Below that, Pro is cheaper. Above, the lower outflow fee on Business pays for the higher subscription. Annual billing pulls the break-even slightly lower (~$850/month outflow). Here is the derivation.

The break-even equation

Pro cost       = 29 + (0.04 x V)   (V = monthly outflow)
Business cost  = 49 + (0.02 x V)

Solve for V where Pro cost = Business cost:

  29 + (0.04 x V) = 49 + (0.02 x V)
        (0.02 x V) = 20
                 V = 1000

Annual billing equivalent rates (~$24 Pro, ~$41 Business per month):
  24 + (0.04 x V) = 41 + (0.02 x V)
        (0.02 x V) = 17
                 V = 850

Below V = $1,000/month outflow per agent: Pro is cheaper. Above: Business is. At exactly $1,000: identical cost ($69 either way on monthly billing).

Worked examples at common volumes

Monthly outflow	Pro total ($29 + 4%)	Business total ($49 + 2%)	Savings on Business
$500	$49	$59	-$10 (Pro wins)
$1,000	$69	$69	$0 (break-even)
$2,000	$109	$89	+$20 (Business wins)
$5,000	$229	$149	+$80
$10,000	$429	$249	+$180
$25,000	$1,029	$549	+$480
$50,000	$2,029	$1,049	+$980

For agents at production scale (think: an MCP server taking thousands of paid invocations daily, an API behind a paywall used by many agents), the math compounds quickly. A single $50k/month agent saves $480/month on Business vs Pro - $5,760/year on one agent, before counting the non-cost features. Below the $2k crossover, Pro remains the right choice unless one of the non-cost triggers applies.

ALLOWLIST DEEP-DIVE

The strictest spend control on the platform.

Daily caps and per-payment caps limit how much an agent can spend. The allowlist limits who it can spend with. Both checks run on every outgoing payment; a payment must satisfy all of them or it is rejected with policy.exceeded.

Two allowlist modes

By wallet address

Pin specific 0x addresses. Outgoing payments to any address not on the list are rejected. Best for: agents paying a fixed set of known vendors (specific MCP servers, specific API providers, specific internal wallets).

allowed_counterparties: [
  "0xAbC...feed",
  "0xDeF...api",
  "0x123...mcp"
]

By verified merchant identity

Pin agent slugs whose domain has been verified by Blockchain0x. Outgoing payments to addresses NOT linked to one of these verified merchants are rejected. Best for: regulated contexts where the destination must be a known business.

allowed_merchants: [
  "research-bot",
  "search-api",
  "translate-mcp"
]

Why allowlist, not blocklist

An allowlist is opt-in: by default, the agent can pay no one. A blocklist is opt-out: by default, the agent can pay anyone except a maintained list of bad actors. We support allowlists, not blocklists. The reason is the threat model: if your agent gets prompt-injected into paying an attacker, the allowlist blocks the attacker regardless of whether you anticipated them. A blocklist only protects against attackers you have already heard of. For agent infrastructure where the prompt-injection surface is large and growing, allowlist is the only model that scales safely.

Operational rules

Allowlist edits are logged in the audit log with the team member who made the change and the list diff.
Edits apply immediately to subsequent evaluations. Pending payments to a now-disallowed address are NOT cancelled.
An empty allowlist means "no allowlist enforcement" - the agent can pay anyone subject to the caps. To block all outgoing payments, set per_payment_cap_usdc to 0.
Allowlists do NOT restrict incoming payments. Anyone can pay your agent at any time regardless of allowlist state.
You can mix modes: use allowed_counterparties for known wallet addresses you trust AND allowed_merchants for verified businesses simultaneously. A payment passes if it matches either list.

AUDIT LOG

Every action, every decision, in a tamper-evident chain.

The Business plan records every action that touches an agent in an append-only log. Each entry references the SHA-256 hash of the previous entry, so retroactive tampering is detectable. Exportable to JSON/CSV/NDJSON; webhookable to your SIEM.

Fields recorded on every entry

Field	Type	Description
audit_id	string	ULID, sortable by time, globally unique.
timestamp	ISO 8601	UTC, microsecond precision, immutable once written.
agent_id	string	Which agent this action affected.
action	enum	policy.evaluated, policy.exceeded, payment.requested, payment.confirmed, settings.updated, member.added, allowlist.updated, key.rotated, plan.changed, etc.
actor_type	enum	human, api_key, system, webhook.
actor_id	string	User ID, API key ID, or 'system' depending on actor_type.
request_ip	string	Source IP for human or api_key actors. Empty for system.
user_agent	string	User-Agent header for HTTP actors. Empty for system.
request_id	string	Correlates with API request logs and webhook delivery records.
before	json	State snapshot before the action (for mutations). Empty for read operations.
after	json	State snapshot after the action. Empty for read operations.
outcome	enum	success, failure, rejected, partial.
reason	string	Human-readable explanation for failure/rejection outcomes.
previous_hash	sha256 hex	Hash of the immediately previous audit entry's full record. First entry is null.

Export and integration

Dashboard export: filter by date range, action, actor, agent. Download as JSON, CSV, or NDJSON.
API: GET /v1/agents/{id}/audit-log with cursor pagination. Supports the same filters.
SIEM webhook: subscribe to the audit.* event family. Every audit write fires a signed webhook to your endpoint within seconds. Compatible with Datadog, Splunk, ELK pipelines.
Retention: 7 years by default on Business. Custom retention windows on negotiated enterprise contracts.

TEAM SEATS

Two kinds of seats, one workspace.

Business introduces a seat model designed for agencies and teams that operate many agents with different operators per agent. Two seat types, with different scopes and prices.

AGENT-LEVEL SEAT

3 included per Business agent

A person who can see and manage ONE specific agent: edit settings, view transaction history, set spend policies, manage allowlists, view that agent's audit log. They cannot see other agents in the workspace, cannot change billing, cannot add or remove team members.

Use case: client operators for an agency. Operator A works on Client X's agent only; Operator B works on Client Y's agent only; neither sees the other.

WORKSPACE SEAT

1 included, $5/seat/month for more

A person with full workspace visibility: every agent, billing, team management, plan changes. The workspace owner is the first workspace seat by default. Add more for senior operators or for finance/ops people who need cross-workspace visibility.

Use case: agency owner, ops lead, finance person who reconciles invoices, security/compliance reviewer.

Adding and removing seats

Workspace owner invites by email from the dashboard. Invitee accepts via email link; account is created (or linked to existing) on first sign-in.
Removing a seat is immediate; the removed person's API keys are revoked atomically with the removal.
Every seat add, role change, and removal is recorded in the audit log with full attribution.
Two-factor authentication is required for workspace seats; agent-level seats can opt in.

FREQUENTLY ASKED

Four Business-plan questions worth answering.

Is Business worth it before my agent crosses $1,000/month in outflow volume?

Often yes, for non-cost reasons. Most teams that upgrade do so because of the recipient allowlist (compliance), the full audit log (regulatory requirements), the higher Spend Permission cap ($100k vs Pro's $10k), or the team-seat sharing (collaboration), not because they did the cost math first. The 2% outflow fee makes the upgrade pay for itself at $1,000/month outflow because of the savings vs Pro's 4%; below that, you are paying for features, not fees. Annual billing pulls the break-even slightly lower (~$850/month outflow). If your team has any of the four feature triggers above, upgrade now rather than waiting for the volume crossover.

How are team seats different from workspace seats?

A workspace seat is a person who can log in and see something in your workspace. By default, Business gives you 1 workspace seat (the owner) plus 3 agent-level seats per Business agent. An agent-level seat is a person who can see and manage one specific agent, without seeing the rest of the workspace. This is useful for agencies: one operator manages Client A's agent, another manages Client B's, neither sees the other. To give one person visibility across the whole workspace, you need workspace seats (add-on, $5/seat/month).

Can I export the audit log for compliance review?

Yes. The audit log is exportable as JSON, CSV, or NDJSON via the dashboard or via GET /v1/agents/{id}/audit-log. Filters by date range, action type, and team-member who took the action. The export includes cryptographic hash chains: each entry references the SHA-256 of the previous entry, so tampering with historical records is detectable. For deeper compliance integration, we also support webhook delivery of audit events to your SIEM (Datadog, Splunk, ELK) via the audit.* event family.

Do I get a discount if I run 10+ Business agents?

Not as a published volume discount, no. The per-agent model deliberately keeps pricing transparent and predictable; we do not negotiate rate cards. If you are running 10+ Business agents and processing meaningful aggregate volume, the conversation usually shifts from per-agent pricing to a true enterprise contract (annual commitment, SOC 2 attestation, MSA, dedicated support). Email [email protected] if you have crossed that scale; we will have a real conversation about what makes sense rather than applying a stock discount.

Run agents at production volume.

Upgrade any agent to Business from the dashboard, or talk to the founder.

Compare plans