Larz API · v1

Build on the Larz ecosystem

One custom, Larz-native API for AI generation and 40+ ready-made tools — plus a registry to publish your own tools into Larz OS. Free to start, no card required.

Quickstart

The Larz API is a clean, Larz-native HTTP API. Base URL:

https://api.larzos.com/v1

Create a key below, then make your first call:

# Generate text
curl https://api.larzos.com/v1/generate \
  -H "Larz-Api-Key: larz_sk_your_key" \
  -H "Content-Type: application/json" \
  -d '{"prompt":"Write a friendly welcome message for a new user"}'

Response:

{
  "id": "larz_gen_82f103691a39…",
  "object": "generation",
  "model": "drlarz-chat",
  "output": "Welcome aboard! We're so glad you're here…",
  "usage": { "tokens": 219 },
  "created": 1783362964
}

Authentication

Every request needs your secret key in the Larz-Api-Key header. (An Authorization: Bearer header is also accepted.) Keys are secret — never expose them in client-side code.

Larz-Api-Key: larz_sk_xxxxxxxxxxxxxxxxxxxxxxxx

Your API keys

Create, name and revoke keys here. Each key starts on the Free tier. Your secret is shown once — store it safely.

Generate

POST /larz/v1/generate

Generate text from a prompt (simple) or a messages array (multi-turn).

Request

FieldTypeNotes
promptstringYour instruction. Use this or messages.
systemstringOptional persona / rules.
messagesarrayOptional. [{role, content}] for multi-turn.
modelstringDefault drlarz-chat. See Models.
max_tokensintClamped to your tier's ceiling.
temperaturefloat0–1. Default 0.7.

Examples

Tools

POST /larz/v1/tools/{slug}
GET /larz/v1/tools

Run any of 40+ ready-made Larz Tools with a single input — no prompt engineering. Browse the catalog with GET /larz/v1/tools.

Models

GET /larz/v1/models
ModelBest forMin tier
drlarz-chatGeneral generationFree
drlarz-fastLow-latency, short tasksFree
drlarz-smartHarder reasoningPro
drlarz-maxHighest qualityPro
drlarz-coderCode generationPro

Usage

GET /larz/v1/usage

Check a key's tier, monthly token usage and limits — handy for showing quota in your own app.

Errors

Errors return an HTTP status plus a Larz error envelope: {"error":{"type","message"}}.

StatustypeMeaning
401authMissing or invalid key
429rate_limitedToo many requests/min
402quota_exceededMonthly token cap reached
403model_not_allowedModel above your tier
503capacityTemporary — retry shortly

Try it live

Paste one of your keys and run a real request from the browser.

Publish a tool

Turn a prompt into a hosted tool that appears in the Larz OS directory and the API — zero hosting. Or register an external endpoint you host yourself (Pro+). Submissions go through a quick review.

Pricing & limits

Free
$0 — no card
  • 60 requests / min
  • 200K tokens / month
  • drlarz-chat & fast
  • Run all 40+ tools
  • Publish hosted tools
Pro
paid
  • 600 requests / min
  • 5M tokens / month
  • All models
  • External plugin tools
  • Priority support
Business
paid
  • 2,000 requests / min
  • 30M tokens / month
  • Pooled team limits
  • Revenue share on tools
  • Priority routing

Every key has a hard monthly token cap, so costs can never run away. Need more? See plans or contact us.