P Paygent

The cost control layer for AI agents.

Meter every LLM call per user per model. Enforce spending limits before they hit your margin. Drop in, ship, done.

$ pip install paygent

The Problem

The hidden cost problem in AI agents

Every agent invocation is a black box of unpredictable spend. Without metering and guardrails, you're flying blind.

Invisible cost multiplication

A single user prompt triggers chains of LLM calls, sub-agents, and tool executions. One prompt. Five model calls. No way to see it happening.

Retries you never see

Your agent retries failed calls and recovers from hallucinations automatically. Great for reliability. Terrible for your bill — because none of it shows up.

Tools blow your budget

Search APIs, code execution, file parsing — every tool call has a cost that varies per user and per session. Forecasting per-user spend is guesswork.

Revenue minus reality

You charge $49/month. Some users cost you $12. Others cost you $140. Same plan, wildly different margins — and you can't tell which is which.

No per-user guardrails

Your heaviest users burn 10x more than they pay for. There's no automated way to throttle, gate, or nudge them toward a higher tier.

Billing flies blind

Your billing system charges a flat monthly fee. Your actual costs vary per user, per session, per prompt. These two worlds never talk to each other.

User Prompt Agent LLM x N Tools x N Rising Cost

What Paygent Does

A runtime layer for controlling AI agent economics.

Three capabilities. One SDK. Full spend control.

01

Metering

Problem

A single user prompt can trigger a chain of model calls, retries, and tool executions. You have no idea what any individual user actually costs you.

Benefit

Track token consumption, tool costs, and agent activity per user, per session, in real time. Know exactly what every invocation costs before it becomes a surprise.

with paygent_context(user_id="user_123"):
user_123 Pro Plan
gpt-4o 31,200 tokens
claude-sonnet 8,400 tokens
gpt-4o-mini 12,100 tokens
Period cost $23.47
02

Guardrails

Problem

You charge every user the same flat fee, but some cost you 10x more than others. Without per-user limits, your best plan subsidizes your worst margins.

Benefit

Set per-user spend limits and per-model token caps. When a user hits their threshold, Paygent automatically triggers a soft warning or hard block — your rules, not manual intervention.

pg.on_soft_gate(on_approaching_limit)
Pro Plan $49/mo
usage 87%
threshold reached
→ trigger upgrade flow
03

Sync & Connect

Problem

Wiring usage data to your backend takes weeks of custom code — aggregation, API endpoints, database queries. It's fragile and it breaks every time you change pricing.

Benefit

Usage events sync to the Paygent backend automatically. Query per-user costs, model-level breakdowns, and billing period summaries via API. Connect the data to your existing billing however you want.

pg = Paygent.init(api_key="pg_live_...")
usage event +142 tokens
billing event $0.021 metered
payment confirmed $17.84
Sync to Paygent backend. Query via API. Connect to any billing system.

Unit Economics Impact

See the difference

Without Paygent

Subscription Revenue $999
Actual AI Cost $1,340
Net Margin -34%

Costs exceed revenue. Margin is negative.

With Paygent

Recommended
Subscription Revenue $999
Controlled AI Cost $719
Net Margin +28%

Guardrails keep costs within budget.

Illustrative example based on typical AI SaaS usage patterns.

Quick Start

Get started in 5 minutes 

from paygent import Paygent, paygent_context

# 1. Initialize (connects to Paygent backend)
pg = Paygent.init(api_key="pg_live_...")

# 2. Wrap your request handler
with paygent_context(user_id="user_123"):
    response = openai.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": query}],
    )

# That's it. Every LLM call is now:
#   ✓ Metered (tokens, cost, per model)
#   ✓ Guarded (soft warnings + hard blocks)
#   ✓ Synced (to Paygent backend, automatically)
Supports OpenAI Anthropic LangChain CrewAI

Ready?

Start metering your AI agent costs today

$ pip install paygent