What is Nolma?

Nolma is an AI gateway platform that sits between your agents and every LLM provider. One URL change gives you:

Real-time cost tracking per agent, model, user, and session
Hard budget enforcement that stops spend before the LLM call fires
Prompt intelligence — track what users do with AI outputs and get recommendations to cut costs

Quickstart

Integrate in 5 minutes

Python SDK

pip install nolma

Node.js SDK

npm install @nolma/node

API Reference

Full API docs

How it works

Your agent code
      |
Change base_url to gateway.nolma.ai
      |
Every LLM call tracked automatically
      |
Dashboard shows real-time costs
Budget enforcement fires if needed

Supported providers

Provider	URL prefix	Models
OpenAI	`/openai`	gpt-4o, gpt-4o-mini
Anthropic	`/anthropic`	claude-sonnet-4-6, claude-haiku-4-5
Google Gemini	`/gemini`	gemini-2.5-pro, gemini-2.0-flash
Groq	`/groq`	llama-3.3-70b
Mistral	`/mistral`	mistral-small-latest

The three products

Guard — Cost control

Real-time cost dashboard. Hard budget enforcement that blocks before the LLM call fires. Zero cost incurred when limits hit.

Lens — Prompt intelligence

Track what users do with AI outputs. Acceptance rates, edit distances, retry patterns. Evidence-backed recommendations to cut costs.

Shield — Security (coming soon)

Injection detection, PII redaction, compliance audit trail. Block threats before they reach the LLM.

Quickstart

​What is Nolma?

Quickstart

Python SDK

Node.js SDK

API Reference

​How it works

​Supported providers

​The three products

​Guard — Cost control

​Lens — Prompt intelligence

​Shield — Security (coming soon)

What is Nolma?

How it works

Supported providers

The three products

Guard — Cost control

Lens — Prompt intelligence

Shield — Security (coming soon)