AI spend accountability infrastructure

Make AI spend
accountable.

P402 turns every AI request into an economic event with owner, workflow, model, provider, tokens, cost, budget, policy result, outcome, and evidence status. Private by default, metadata-only metering.

No prompt storage required. Metadata-only mode is available from the first event.

The problem

AI spend arrives too late.

Provider invoices show totals after usage happens. They do not show which customer, feature, workflow, department, employee, model, or retry caused the cost. P402 records that ownership when the event happens.

Nov 2026in flight
Without P402
no owner attached
Nov 1
Nov 30
pending
$ ___,___.__
one row
no owner
With P402
event ledger live
Nov 1
Nov 30
live total
$0.00
0 of 24 events
Install in your environment
VS Code / Cursor / Windsurf
ext install p402-protocol.p402

Embedded MCP server. Tools appear in Copilot agent mode on install. No config files.

Claude Desktop / JetBrains
npx -y @p402/mcp-server

stdio MCP server. Add to your client config with P402_API_KEY.

REST API / SDK
npm install @p402/sdk

OpenAI-compatible endpoint. Drop-in replacement. Change the base URL and API key.

Trust

What P402 commits to.

Privacy boundary

Meter economics, not content. Metadata-only by default. Five privacy modes from metadata-only to full-trace opt-in.

Status boundary

Meter is shipped. Control runtime flip remains gated. Optimize recommendations remain gated until outcome data is sufficient.

Proof object

Dashboard event detail, evidence bundle, finance report, CSV appendix.

Conversion next step

Start free, Run AI Spend Audit, See evidence, Apply as partner, or Read docs.

Live Routing

Every request.
Routed live.

P402 selects the optimal model for each call. You write the same OpenAI-compatible code.

Incoming Request
Mode:Cost
prompt
Summarize this legal contract in plain English.
Modelclaude-opus-4-6deepseek-chat
Cost$0.01500$0.00041(97% saved)
p402_metadata
● routing...
// selecting optimal provider...
300+ models·Cost / Speed / Quality / Balanced·OpenAI-compatible·Full audit trail·Semantic cache

Transparent Economics

Free Sandbox. Production plans usage-based. Enterprise sales-led.

Evaluate

Sandbox

$0
/month

25,000 metered events. Hard cap.

Start free
Build

Build

$49
/month

250,000 events. 30-day retention.

Start Build
Operate

Growth

$199
/month

2,000,000 events. 90-day retention.

Start Growth
Production Ready

Real Infrastructure
For Real Payments

P402 is live in production with real USDC settlements on Base mainnet. No testnets, no simulations - just working payment infrastructure.

Gasless USDC Payments

Live on Base Mainnet

Users pay zero gas fees with EIP-3009 transferWithAuthorization

100%
Gas Savings
<2 seconds
Settlement Time
1%
Fee Rate

Global Facilitator Network

Production on Cloudflare

Edge deployment across 15 regions with sub-50ms verification

15
Global Regions
<50ms
P95 Latency
99.9%
Uptime SLA

Real-Time Cost Tracking

Dashboard Available

Live cost monitoring, transaction history, and savings analytics

5 seconds
Update Frequency
100%
Cost Transparency
Historical Data

Multisig Treasury Security

Security Hardened

3-of-5 signature requirement with emergency pause capabilities

3 of 5
Required Signatures
Active
Emergency Controls
Complete
Audit Trail

Gasless USDC Payments

Users pay zero gas fees with EIP-3009 transferWithAuthorization

// EIP-3009 Flow
1. User signs authorization (no gas)
2. P402 verifies signature
3. Facilitator executes transfer
4. User receives service instantly
Try Live Demo

Production Infrastructure

$0
Gas Fees for Users
15
Global Edge Regions
99.9%
Uptime Guarantee
3/5
Multisig Security

Built on an open standard

P402 implements Google Cloud's open Agent-to-Agent (A2A) protocol.

The quotes below are public statements by partners of the A2A standard about the protocol itself, not endorsements of P402. Source: Google Cloud A2A launch announcements.

"Ask-AI is excited to collaborate with Google on the A2A protocol, shaping the future of AI interoperability."

Ask-AI
CEO Alon Talmor PhD
Statement about A2A protocol

"A standardized protocol like A2A will help agents successfully discover, coordinate, and reason with one another to enable richer forms of delegation."

Atlassian
Brendan Haire, VP Engineering AI Platform
Statement about A2A protocol

"We’re enabling Articul8's ModelMesh to treat A2A as a first-class citizen, enabling secure, seamless communication between intelligent agents."

Articul8
Arun Subramaniyan, CEO
Statement about A2A protocol

"We look forward to enabling Box agents to work with Google Cloud’s agent ecosystem using A2A... empowering organizations to better automate workflows."

Box
Ketan Kittur, VP Product Management
Statement about A2A protocol

"A2A has the potential to help customers break down silos and securely enable AI agents to work together across systems, teams, and applications."

C3 AI
Nikhil Krishnan, CTO Data Science
Statement about A2A protocol

"As a pioneer in enterprise multi-agent systems, Cognizant is committed and actively pursuing agent interoperability as a critical requirement."

Cognizant
Babak Hodjat, CTO - AI
Statement about A2A protocol

"The open A2A protocol ensures seamless, trusted collaboration—even in air-gapped environments—so that businesses can innovate at scale."

Cohere
Autumn Moulder, VP Engineering
Statement about A2A protocol

"We're excited to see Google Cloud introduce the A2A protocol to streamline the development of sophisticated agentic systems."

Datadog
Yrieix Garnier, VP Product
Statement about A2A protocol

"Salesforce is leading with A2A standard support to extend our open platform, enabling AI agents to work together seamlessly."

Salesforce
Gary Lerhaupt, VP Product Architecture
Statement about A2A protocol

"ServiceNow and Google Cloud are collaborating to set a new industry standard for agent-to-agent interoperability."

ServiceNow
Pat Casey, CTO
Statement about A2A protocol

"Establishing an industry standard for seamless agent-to-agent communication is a significant step towards a future in which AI agents and humans collaborate."

UiPath
Graham Sheldon, CPO
Statement about A2A protocol

"Combining MongoDB with A2A lets businesses unlock new possibilities across industries to redefine the future of AI applications."

MongoDB
Andrew Davidson, SVP Products
Statement about A2A protocol

"PayPal supports Google Cloud’s A2A protocol, which represents a new way for developers and merchants to create next-generation commerce experiences."

PayPal
Prakhar Mehrotra, SVP AI
Statement about A2A protocol

A distribution loop tied to discovery.

Bazaar is the discovery layer for payable endpoints. Router adds runtime signals so buyers pick endpoints using real health and policy outcomes.

Listing metadataRuntime healthPolicy outcomes
Explore the Bazaar →
Bazaar Overlay Preview
Approval Rate
98.2%
p95 Verify
42ms
p95 Settle
310ms
Last Seen
12s ago
exact, USDC, eip155:8453
Meter is shipped.

Give every token
an owner.

Start with one metered event.

P402 Meter | Make AI Spend Accountable