AI Cost Observability + Guardrail Infrastructure

Know exactly what your
AI agents are costing you.

Deep cost attribution per workflow, step, and customer — plus hard budget enforcement that actually stops agents when limits hit.

Be first to know when we launch · No spam · Ever
Caltryx. — production workspace
Live
Dashboard
🤖 Agents
🔄 Workflows
👥 Customers
🔍 Tracing
🛡️ Budgets
📱 Telegram
🔑 API Keys
Total Cost Today
$142.80
↑ 12% vs yesterday
Active Agents
12
3 workflows running
Budget Used
87%
Tokens Today
4.2M
↓ 8% efficient
Cost per Workflow today
support
data_sync
onboard
analyst
summarize
$42$74$28$54$18
Top Customers by cost
acme_corp
$200
startup_x
$41
user_4821
$22
dev_team
$11
⚠️ acme_corp nearing monthly budget cap
Built for
Solo SaaS founders building with LLMs· Devs shipping AI features· Indie hackers· Small AI startups· Multi-agent builders· B2B SaaS embedding AI· Solo SaaS founders building with LLMs· Devs shipping AI features· Indie hackers· Small AI startups· Multi-agent builders· B2B SaaS embedding AI·

One SDK. Total
visibility.

Three steps from zero visibility to full cost attribution across every agent, workflow, and customer.

🔌
01

Drop in the SDK

Install the Caltryx SDK in minutes. Wrap your LLM calls with trackLLM() and attach tags like user_id, workflow_id, and step_name. That's it.

⚡ 5-minute integration
02

Tags power everything

Every token, every call, every dollar gets attributed. Tag by tenant, workflow, step, model, environment — or any custom dimension. Without tags, nothing works. So we make tagging dead simple.

🎯 Unlimited custom dimensions
03

Set budgets, enforce hard limits

Not just alerts — hard enforcement. Set daily, monthly, per-workflow, per-tenant, or per-agent budgets. Caltryx auto-stops agents and blocks API calls the moment a limit is hit.

🔒 Kill switch included

You're flying blind on
AI costs. That ends now.

Most AI builders don't know what they're spending — until the invoice arrives and it's too late.

💸
No idea which workflow costs the most
You have 5 agents running. You have no idea if support_bot is $4 or $400 a day. You find out at month end.
🏢
Can't attribute costs to customers
Your enterprise client uses 50x more AI than others. You're billing them the same flat fee. You're losing money.
🚨
No hard budget enforcement
You set an alert. The agent ignores it. You get an email after $2,000 of runaway spend. Alerts aren't guardrails.
🕳️
Unknown spend from unmapped traffic
Untagged calls. Missing tenant IDs. Rogue agents. A significant chunk of your AI spend is completely invisible.
🔍
Zero step-level tracing
A multi-step pipeline costs $18. You don't know if it's the planner, the executor, or the validator burning budget.
📊
No model optimization signal
GPT-4 doing work GPT-3.5 could handle. Large contexts where small ones would do. You're over-paying by default.

Every layer of AI cost
infrastructure, covered.

From agent management to Telegram control — Caltryx is the complete observability stack for production AI.

01
Agent Management System
Create, edit, enable/disable agents. Assign models, temperature, max tokens, and API keys per agent.
02
BYOK System
Add your own OpenAI or Anthropic keys. Encrypted storage, per-agent assignment, per-key usage tracking.
03
Deep Cost Attribution
Cost per workflow, per step, per customer. Powered by the tagging layer — the strongest value in the stack.
04
Guardrails & Budget Control
Hard enforcement — not just alerts. Kill switch, auto-stop, API blocking, throttle, and manual override.
05
Observability & Tracing
DevOps-style tracing with Correlation IDs. Full request logs, error logs, token breakdowns, response times.
06–09
Cost Optimization · Telegram Bot · SDK · REST API
Token waste detection, model swap suggestions, Telegram control center, drop-in SDK, webhooks, and CSV export.
Cost Attribution — Live
🔄
support_bot workflow → $74 today
Planner $4 · Executor $9 · Validator $1 · per step
👤
acme_corp (tenant) → $200 this month
Highest-cost customer — profitability at risk
🚨
Budget $500/mo — 87% used
Kill switch armed — will auto-stop at limit
Optimization signal
Switch summarizer to GPT-3.5 → save $31/day
SDK — drop-in tracking
trackLLM({ workflow: "support",
  step: "planner", user_id: "u123",
  tenant_id: "acme_corp" })

Infrastructure pricing.
Not toy pricing.

Start solo. Scale to enterprise. Every plan pays for itself the first time it stops a runaway bill.

Active
Starter
$19
/ month (USD)
$228 billed yearly
For solo AI builders getting started. Essential visibility on 5 agents.
5 agents
5M tracked tokens / month
External tracking mode
Budget alerts
SDK mode
Slack / Telegram alerts
Webhooks + CSV export
Advanced analytics
Save 27%
Growth
$65
$49
/ month (USD)
$588 billed yearly
For serious AI startups that need step-level attribution, SDK, and hard enforcement.
20 agents
25M tracked tokens / month
External + SDK mode
Slack / Telegram alerts
Webhooks + CSV export
Advanced analytics
Priority email support
Multi-user workspace
Scale
$129
/ month (USD)
$1,548 billed yearly
For AI-heavy teams needing multi-user workspaces, role permissions, and full control.
75 agents
100M tracked tokens / month
Multi-user workspace
Role-based permissions
Custom alert rules
Everything in Growth
Faster support
Dedicated infrastructure

🏛 Enterprise

Built for AI-native companies operating at scale. Custom pricing, on request.

Unlimited agents + tokens
Dedicated infrastructure
SLA agreement
White-label option
Custom integrations
Dedicated support + onboarding

All plans include BYOK · No hidden fees · Cancel anytime

Caltryx is
AI cost infrastructure.

Not an analytics dashboard. Not an APM tool. The observability + guardrail layer built for LLM workloads.

❌ Caltryx is NOT
A generic APM / monitoring tool
A Datadog replacement
An LLM playground
A prompt management tool
An alert-only budget tool
A replacement for your AI stack
✓ Caltryx IS
Deep cost attribution per tenant
Hard budget enforcement + kill switch
Step-level workflow tracing
BYOK with encrypted key storage
Drop-in SDK + Telegram control
Agent management + cost optimization

Infrastructure for serious AI builders.

0
Complete feature architecture out of the box
0
SDK integration time — start tracking immediately
0
Hard budget enforcement — not just alerts
BYOK
Bring your own API keys — always encrypted

Your questions, answered

Everything before you integrate Caltryx into production.

What exactly is Caltryx? +
Caltryx is AI Cost Observability + Guardrail Infrastructure combined with an Agent Management Layer. It gives you deep cost attribution per workflow, per step, and per customer — plus hard budget enforcement that actually stops agents when limits are hit. Not just alerts.
How does BYOK (Bring Your Own Key) work? +
You add your OpenAI or Anthropic API keys to Caltryx — they're encrypted at rest and never stored in plain text. You can assign different keys per agent, track usage per key, and manage multiple keys per workspace. Caltryx never charges you for LLM calls — you pay your providers directly.
What's the difference between alerts and hard budget enforcement? +
Most tools send an email after you've overspent. Caltryx enforces budgets in real-time — auto-stopping agents, blocking API calls, or throttling usage the moment a limit is reached. Manual override is always available from the dashboard or directly via Telegram command.
How does tenant-level cost attribution work? +
When you call trackLLM(), you attach a tenant_id tag. Caltryx aggregates all token usage and cost against that tag in real-time. You see exactly which customers are profitable, which are power-users burning budget, and how to accurately price your AI-powered features.
Does Caltryx support multi-agent and LangChain workflows? +
Yes. Caltryx is designed specifically for multi-agent workflows. Correlation IDs link every step of a chain into a single trace — so you see the full cost breakdown across every node (planner, executor, validator) in one view, with per-step attribution.

Stop guessing.
Start knowing.

Join the waitlist. Be first when Caltryx launches — and help shape exactly what we build.

Be first to know when we launch · No spam · Ever

Everything you need to run AI in production.

Caltryx is not one feature — it's the complete observability + guardrail stack. Every system works independently, and powerfully together.

01

Agent Management

Create, edit, enable/disable agents. Set model, temperature, max tokens, and API key per agent. Full lifecycle control from a single interface.

02

BYOK System

Add your own OpenAI or Anthropic keys. Encrypted at rest. Assign per agent. Track usage per key. Multiple keys per workspace supported.

03

Deep Cost Attribution

Cost per workflow, per step, per customer. The most powerful dimension in the stack — powered entirely by the tagging layer. Every dollar tracked.

04

Guardrails & Budget Control

Hard enforcement — not just alerts. Set daily, monthly, per-workflow, per-tenant, per-agent budgets. Kill switch auto-stops agents when limits hit.

05

Observability & Tracing

DevOps-style tracing with Correlation IDs. Full request logs, error logs, model used, response time, and complete token breakdowns per step.

06

Cost Optimization Layer

Token waste detection. Large context warnings. Model swap suggestions. Unknown spend detection for untagged traffic. Anomaly detection on usage spikes.

07

Telegram Control Center

Alerts for budget exceeded, usage spikes, agent stopped. Run /stats, /usage, /stop agent-name, /budget directly from Telegram. Lightweight remote control.

08

Drop-in SDK

Node/JS first. One function call wraps your LLM. Attach tags, report tokens, enforce budgets, generate correlation IDs. 5-minute integration guaranteed.

09

REST API + Webhooks

Full REST endpoint. Webhook support for real-time events. Usage and CSV export. Team access with per-workspace API keys. Build on top of Caltryx.

Caltryx is launching soon. Join the waitlist — be first to access all 9 systems.

From zero visibility to full
cost clarity in 5 minutes.

Three steps. One SDK. Total attribution across every agent, workflow, and customer in your stack.

01

Install the SDK

One npm install. Works with any Node/JS codebase — LangChain, raw OpenAI calls, custom agents, multi-step pipelines. No lock-in.

Terminal
npm install @Caltryx/sdk
02

Wrap your LLM calls with trackLLM()

One function call does everything: attaches tags, reports tokens, enforces budgets, and generates a correlation ID for full trace linking.

your-agent.js
import { trackLLM } from '@Caltryx/sdk';

await trackLLM({
  workflow: "support",
  step: "planner",
  user_id: "u123",
  tenant_id: "acme_corp",
  model: "gpt-4o"
});
03

Set budgets. Enforce hard limits.

From the dashboard or API: set daily/monthly budgets per agent, workflow, or tenant. Caltryx enforces in real-time — auto-stop, block, or throttle. Telegram alert fires instantly.

Dashboard or API
POST /budgets
{ tenant_id: "acme_corp", monthly: 500,
  action: "stop" }
Ready to integrate? Join the waitlist and be first when the SDK ships.

Start solo.
Scale to enterprise.

Every plan includes BYOK. Every plan pays for itself the moment it catches one runaway LLM bill.

🧱
Starter
$19
per month
5 agents
5M tokens / month
External tracking mode
Budget alerts
Basic analytics
Email notifications + API
🏢
Scale
$129
per month
75 agents
100M tokens / month
Multi-user workspace
Role-based permissions
Custom alert rules
Everything in Growth
🏛
Enterprise
Custom
on request
Unlimited agents + tokens
Dedicated infrastructure
SLA + audit logs
White-label option
Custom integrations
Dedicated support + onboarding

All plans include BYOK · No hidden fees · Cancel anytime

Pricing locked in for early members. Join the waitlist before launch.

Dead simple.
5 minutes to full visibility.

The Caltryx SDK is Node/JS first. One function wraps your LLM calls and handles everything — tags, tokens, budgets, correlation IDs.

Installation

npm install @Caltryx/sdk

Basic usage

import { trackLLM } from '@Caltryx/sdk';
import OpenAI from 'openai';

const client = new OpenAI();

const result = await trackLLM({
  workflow: "support-bot",
  step: "planner",
  user_id: "user_123",
  tenant_id: "acme_corp",
  model: "gpt-4o",
  fn: () => client.chat.completions.create({ ... })
});

Supported tag fields

workflowGroup calls by workflow name
stepIdentify the step within a workflow
user_idLink cost to an individual user
tenant_idB2B tenant attribution
workspace_idMulti-workspace separation
modelOverride model tracking
environmentprod / staging / dev
custom.*Any custom key-value pair

Budget enforcement

// Set in dashboard or via API
POST /v1/budgets
{
  "tenant_id": "acme_corp",
  "monthly_usd": 500,
  "action": "stop" // or "throttle" / "alert"
}

Everything you need
to build with Caltryx.

Full reference for the Caltryx SDK, REST API, webhook events, and dashboard configuration.

Thinking about AI
cost infrastructure.

Practical writing for AI builders who care about what they're spending.

Cost Attribution

Why your AI bill will surprise you — and how to stop it

Most founders discover their AI costs through the invoice. Here's how to see it coming before it destroys your margins.

5 min read · Feb 2026
Guardrails

Alerts vs. hard limits: why one actually protects you

An alert tells you after the damage. A kill switch prevents it. The distinction matters more than most people realise.

4 min read · Feb 2026
BYOK

The right way to manage API keys across multiple agents

One key for everything is a security and attribution disaster. Here's the architecture that actually scales.

6 min read · Jan 2026
Multi-agent

Tracing cost across a 5-step LangChain pipeline

Correlation IDs are the key to understanding where your money goes inside complex agent workflows.

8 min read · Jan 2026

What we've shipped.

Every update, improvement, and fix — in reverse chronological order.

Feb 2026
New

Waitlist launched

Caltryx is officially in pre-launch. Join the waitlist to be first when we open access. Early waitlist members get 3 months of Growth free.

Coming soon
Soon

SDK v1 — Node/JS

The drop-in trackLLM() SDK for Node and JS environments. BYOK support, full tagging, budget hooks, and correlation IDs built in.

Coming soon
Soon

Dashboard beta

The full Caltryx dashboard — cost per workflow, per step, per customer. Agent management, budget controls, tracing view, and Telegram integration.

Planned
Planned

Anomaly detection + optimization layer

Token waste detection, model swap recommendations, and unknown spend flagging powered by internal analysis.

Your keys.
Always encrypted.

Caltryx handles API keys, usage data, and cost attribution with security-first architecture.

🔑

BYOK Encryption

All API keys are encrypted at rest using AES-256. Keys are never stored in plain text, never logged, and never transmitted without encryption.

🛡️

No LLM data stored

Caltryx tracks tokens, costs, and metadata — not your prompts or completions. We never store or see the content of your LLM calls.

🔒

Workspace isolation

All data is scoped to your workspace. No cross-tenant data access. Role-based permissions on Scale and Enterprise plans.

📋

Audit logs

Complete audit trail for all budget changes, agent modifications, API key additions, and permission changes. Available on Enterprise plans.

Built by an AI builder
who got burned.

Caltryx was born from a real problem: a multi-agent pipeline quietly running up a $600 bill with zero visibility into which step, workflow, or customer was responsible.

The existing tools weren't built for LLM workloads. APMs track infrastructure. Analytics dashboards show page views. Neither tells you that your summarizer agent is spending $40/day when a cheaper model would do the same job for $4.

Caltryx is infrastructure for the way AI is actually built in 2026 — multi-agent, multi-tenant, multi-model, and increasingly expensive if you're not watching it closely.

We believe every indie hacker, solo founder, and small AI startup deserves the same cost visibility and guardrail tools that enterprise companies build internally. That's what we're building.

Caltryx is currently in pre-launch. Join the waitlist — early members help shape what we prioritise.

🎯
Focused

We do one thing: AI cost observability and budget enforcement. No scope creep.

🔌
Drop-in

5 minutes to full visibility. No rearchitecting. No lock-in.

💬
Builder-first

Built for indie hackers and solo founders — not enterprise IT departments.

Join us from day one. Early waitlist members get 3 months of Growth free.

Let's talk
AI infrastructure.

The best way to reach me is a DM on X. Questions, enterprise inquiries, partnership ideas, or just your LLM cost horror story — I read everything.

𝕏
@showslikesummu

DM me directly on X. I reply to every genuine message. Building Caltryx in public — feedback, feature requests, and ideas all welcome.

Open X.com / DM me →
🏢
Enterprise inquiries

DM on X and mention "Enterprise" — I'll schedule a call same week.

💡
Feature requests

Building for builders. Every idea goes into the roadmap. DM or quote-tweet.

🐛
Bug reports

Found something wrong? DM with details. Fast fixes, always.

Caltryx is launching soon. Join the waitlist — be first in the door.

Your data.
Handled with care.

Last updated: February 2026

Simple, fair terms.

Last updated: February 2026