Demeterics AI Platform
Demeterics AI Platform

Precision Instrumentation for AI

Tag Every LLM Call.
Slice Data Any Way.

Add /// FLOW checkout to your prompts. Instantly see costs by app, workflow, market, variant, and more. No SDK. No code changes. Just comments.

Start Tracking Free See How Tags Work

100 free credits included. Works with OpenAI, Anthropic, Groq, and 50+ providers.

5-Minute Setup

Zero Code Changes. Instant Visibility.

Route your AI calls through our proxy. Add smart tags to your prompts. Get granular analytics across every dimension that matters to your business.

1
Change One URL

Replace api.openai.com with api.demeterics.com. Same SDK, same code, same API keys.

2
Add Smart Tags

Add /// APP, /// FLOW, /// MARKET to your prompts. We extract dimensions automatically.

3
Slice Any Way

Filter and aggregate by any dimension. Compare A/B variants. Track costs by team, market, or customer.

Example: Tagged Prompt in Any Tool Works in n8n, Zapier, Make, REST
/// APP customer-support
/// FLOW ticket-response
/// MARKET enterprise
/// VARIANT v3-empathetic

You are a helpful customer support agent.
Respond to the following ticket with empathy...
View Tagging Guide

Our Secret Weapon

Instrumentation Tags: Precision Without Friction

Other tools require SDKs, code wrappers, or external metadata APIs. Demeterics uses inline comment tags that work everywhere— inside prompts themselves, invisible to the LLM, but captured by our proxy.

/// APP

Application or service name

/// FLOW

Business workflow or process

/// MARKET

Geographic or segment

/// VARIANT

A/B test or experiment

/// VERSION

Prompt template version

/// UNIT

Team or department

/// COMPANY

Customer or client ID

/// SESSION

Conversation thread

Works Everywhere

n8n nodes, Zapier Zaps, Make scenarios, raw curl commands. If you can edit a prompt, you can tag it.

Invisible to the LLM

Tags are stripped before reaching the model. Zero extra tokens. Zero impact on output quality.

Query Any Dimension

Slice by app, flow, market, variant—or any combination. Filter, compare, and aggregate instantly.

The more you tag, the more valuable your data becomes. Historical analytics compound over time.

Start Tagging for Free

Real-World Examples

See How Teams Use Smart Tags

From startups to enterprises, teams use instrumentation tags to unlock insights that were previously impossible to capture.

A/B Test Prompt Variants

Compare conversion rates between different prompt strategies. Know which approach works best before rolling out to all users.

/// FLOW onboarding-welcome
/// VARIANT friendly-casual
Welcome! I'm here to help you get started...

Filter dashboard by VARIANT to see cost, latency, and quality per test group.

Track Cost Per Customer

Know exactly how much each customer costs in LLM spend. Bill accurately for AI-powered features or identify abuse patterns.

/// COMPANY acme-corp
/// UNIT sales-team
Summarize the following sales call transcript...

Export COMPANY-level spend for invoicing or margin analysis.

Analyze by Market Segment

Compare AI performance across regions, languages, or customer segments. Optimize prompts for each market.

/// MARKET latam-spanish
/// FLOW product-recommendation
Recomienda productos basados en...

See if Spanish prompts cost more or perform differently than English.

Track Prompt Versions

Know which prompt version is running in production. Measure regressions after updates. Roll back with confidence.

/// APP email-generator
/// VERSION 2.3.1
Write a professional email that...

Compare VERSION 2.3.1 vs 2.2.0 latency, cost, and error rates.

Platform Features

Everything You Need to Optimize AI Spend

Tags unlock the data. These features help you act on it.

Cost Analytics

Track spend by APP, FLOW, COMPANY, or any tag. No more surprise bills.

Latency Tracking

Average, P90, P99 by workflow. Compare VARIANT performance.

Error Debugging

Full request/response logs. Filter errors by any tag dimension.

Dimension Aggregation

Group by any tag. Pivot across MARKET, UNIT, or VERSION.

Multi-Provider

OpenAI, Anthropic, Groq, Gemini, OpenRouter. One dashboard for all.

Token Analysis

Input/output breakdown. Find which prompts waste tokens.

Session Tracking

Link calls by SESSION. Replay full conversation threads.

REST API + Exports

Query data programmatically. Export to CSV or JSON for analysis.

Why Demeterics

The Only Observability That Works Everywhere

Other tools require SDKs, code changes, or complex integrations. Our inline tagging approach means you can instrument any prompt, anywhere.

Other Tools
  • Require SDK integration
  • Code changes needed
  • Don't work in no-code tools
  • Limited metadata options
Demeterics
  • No SDK required
  • Just add comments
  • Works in n8n, Zapier, Make
  • 14+ metadata dimensions
5-Minute Setup

Change one URL.
Add tags to prompts.
Done.

Data Compounds

Historical analytics grow more valuable over time.
Your data is your moat.

Works with all major LLM providers

OpenAI
Anthropic
Groq
Gemini
Mistral
OpenRouter
xAI
Cohere

Transparent Pricing

Pay only for what you use. No hidden fees.

$0

to start

100 credits

included free

~$0.001/call

typical proxy cost

You pay LLM providers directly. We charge only for tracking and analytics.

Start in 5 minutes

Ready to See Every Dimension of Your AI Spend?

Change one URL. Add a few /// tags to your prompts. Unlock insights that grow more valuable every day.

Start Tagging Free Read the Tagging Guide

100 free credits included. | No credit card required. | Works with 50+ LLM providers.