Python 3.11 · Sandboxed · Runs forever

BETA Official CLI on PyPI: pip install litelambda-cli → NEW Model Context Protocol (MCP) for Claude, Cursor & AI →

Run your Python in the cloud.
On a schedule, on a webhook, or from your AI agent.

Every run is isolated in an ephemeral sandbox with state, logs, and failure alerts. Deploy from your terminal, browser, or directly inside Claude, Cursor, and Antigravity.

Trusted by 500+ developers running automated scripts, scrapers, bots, and AI agents worldwide.

Start today — no card needed → See pricing (from $4.99/mo)

15-day trial included · No credit card required · GDPR compliant

Python 3.11 Sandbox

import requests

from litelambda import kv, notify

def handler(event, context):

# 1. Fetch current price without managing a Linux server

res = requests.get("https://api.coinbase.com/v2/prices/BTC-USD/spot")

current_price = float(res.json()["data"]["amount"])

# 2. Compare against previous run with built-in KV store

last_price = kv.get("btc_last_price", default=current_price)

delta = ((current_price - last_price) / last_price) * 100

# 3. Alert to Slack or Discord only on meaningful price drop

if delta <= -2.0:

notify.slack(f"⚠️ BTC Price Drop: ${current_price:,.2f} ({delta:.1f}%)")

# 4. Save state for next hourly run (no PostgreSQL or Redis needed)

kv.set("btc_last_price", current_price)

return {"status": "ok", "price": current_price, "delta": delta}

import stripe, os

from datetime import datetime, timedelta

from litelambda import notify

def handler(event, context):

stripe.api_key = os.environ["STRIPE_SECRET_KEY"]

# Calculate yesterday's net volume in 2 lines

yesterday = int((datetime.now() - timedelta(days=1)).timestamp())

charges = stripe.Charge.list(created={"gte": yesterday}, limit=100)

net_revenue = sum(c.amount for c in charges.auto_paging_iter() if c.paid) / 100

# Dispatch formatted financial digest to private revenue channel

notify.slack(

channel="#revenue",

message=f"💰 *Daily Stripe Digest*: ${net_revenue:,.2f} collected yesterday!"

)

return {"net_revenue": net_revenue, "status": "dispatched"}

# In Claude, Cursor, ChatGPT, or Antigravity, simply ask:

# "Deploy an hourly health check that alerts Telegram if our API is down."

from litelambda import notify

import requests

def handler(event, context):

try:

res = requests.get("https://api.example.com/health", timeout=5)

res.raise_for_status()

return {"status": "healthy", "latency_ms": res.elapsed.total_seconds() * 1000}

except Exception as e:

notify.telegram(f"🚨 API Health Check Failed: {str(e)}")

raise

200 OK | 142ms | 18MB RAM | KV state saved

Deploy your Python script →

Beyond Cron • Four Triggers, One Ephemeral Engine

How Developers Run Python on LiteLambda

From autonomous AI agents to payment webhooks and persistent watchers, every workload runs isolated in a micro-sandbox.

Agent Execution MCP

AI Coding Agents

Let Claude, Cursor, ChatGPT, and Antigravity deploy scripts, inspect logs, and manage state in the cloud without local terminals.

Explore Agent Layer →

Webhooks & Bots Instant URL

Webhook Backends

Paste a Python handler, get an HTTPS URL. Process Razorpay payments, GitHub events, and Telegram bots with built-in idempotency.

Explore Webhooks →

Persistent Watchers Built-in KV

State-Aware Monitors

Compare current values against previous runs with persistent Key-Value storage. Alert to Slack, Discord, or Telegram only on deltas.

Explore Monitors →

Synchronous API ?sync=true

Python Function → API

Turn any Python script into a synchronous JSON API endpoint. Returns the script's dictionary result directly in the HTTP body.

See Sync API Docs →

Custom Agent Tools MCP Tool

Custom MCP Tools

Turn any Python script into a custom AI agent tool. Claude, Cursor, and Antigravity discover and execute your functions in an isolated cloud sandbox.

Explore Custom Tools →

Infrastructure Early Access

Sandbox-as-an-API

Execute untrusted, third-party Python code inside hardened gVisor micro-containers with memory caps and strict network firewalls.

Request Waitlist Access →

New Feature • Model Context Protocol (MCP)

The Background Execution Cloud
for AI Coding Agents.

When you ask your AI assistant to build a recurring price scraper, an automated digest, or a database sync, it shouldn't get trapped running on your laptop. LiteLambda gives Claude, Cursor, ChatGPT, and Antigravity a sandboxed cloud environment to deploy, schedule, test, and monitor Python cron jobs autonomously.

mcp.litelambda.in • Streamable HTTP + SSE

Live Agent Bridge

AI Coding Agent

Claude • Cursor • Antigravity • ChatGPT

MCP Client

Developer Prompt:

"Create an hourly script that monitors competitor pricing, writes low prices to KV, and alerts Slack if a price drops >10%."

Agent Reasoning... Auto-Detected

Generated monitor_prices.py (Python 3.11). Decided schedule: 0 * * * *.

tool: deploy_job

JSON-RPC 2.0

Streamable
HTTP

↓ MCP Tool Call Transmission ↓

LiteLambda Cloud Sandbox

https://mcp.litelambda.in/mcp

Execution Host

[00:00:01] MCP POST /mcp • Bearer Auth ✓

[00:00:01] Tool 'deploy_job' called by agent

[00:00:02] Container sandbox spin-up (184ms)

[00:00:02] pip cached: requests, beautifulsoup4

[00:00:03] KV partition attached: context.kv

✓ Job active • Schedule: 0 * * * *

[00:00:03] Next run in 59m • Webhook generated

Ephemeral Container Provisioned Zero Local Resources

Agent response back to developer: “I've deployed your price scraper to LiteLambda! It is scheduled for hourly runs with Slack alerting and persistent state.”

See MCP Protocol Specs →

Supported AI Coding Environments

Connect in one click or one line of configuration.

Claude.ai • Code

Anthropic Claude

Add LiteLambda as a Custom Connector in Claude.ai or register into Claude Code CLI with a single command. Manage crons in natural language.

Guide →

1-Click Install

Cursor IDE

Deep-link integration into Cursor Composer and Chat Agent. Ask Cursor to deploy Python background routines or inspect production logs without switching windows.

SSE Protocol

Google Antigravity

Native Server-Sent Events (SSE) support for Antigravity IDE & Gemini Agents via mcp_config.json. Launch autonomous multi-agent background schedules.

Streamable SSE

Custom GPTs

OpenAI ChatGPT

Connect LiteLambda's HTTP endpoint as a Custom GPT Action or Assistant tool. Enable ChatGPT to deploy and execute Python scripts on scheduled crons directly from chat prompts.

Streamable HTTP

Ready to empower your AI coding agents?

Zero infrastructure setup. Hosted endpoint is active and free for all LiteLambda users.

Instant Webhook Triggers • On-Demand

Scheduled or on-demand.
Trigger from anywhere.

Need to run your script right after a Stripe payment, GitHub push, or Zapier event? Every cron job gets a secure instant webhook URL. Run on your regular schedule, or trigger in real-time with zero cold starts.

POST External Webhook Call

# Trigger from terminal, Stripe, or CI

curl -X POST https://litelambda.in/crons/webhook/ec98161c/... \

-H "Content-Type: application/json" \

-d '{"event": "invoice.paid", "amount": 9900}'

Auth: UUID Secret in Path Zero Cold Start

function.py • Execution 200 OK • 342ms

def handler(event, context):

# Payload automatically injected

payload = event.get('body', {})

print(f"Processing {payload.get('event')}")

# [INFO] Processing invoice.paid

return {"status": "success"}

Native JSON response returned Docs →

How It Works

Deploy a cron job in under a minute

Write Your Code

Paste your Python script and list any pip packages you need. We handle the environment.

Set a Schedule

Use standard cron syntax — 0 9 * * * for every day at 9 AM. UTC-based. No guesswork.

Go Live

Hit activate. We run it on schedule, store logs, and alert you if anything goes wrong.

Built-in KV Store • BETA

Your job needs to remember things.
Now it can.

Every cron job gets a built-in key-value store — no separate database or Redis cluster to provision. Track what you've already sent, resume where you left off, and count anything across runs.

✓ Zero configuration: Available immediately via context.kv
✓ State persists between runs: Up to 100 keys per job, 64KB per value
✓ Live dashboard view: Inspect and debug your stored values in the editor

Try LiteLambda with KV Store →

sync_orders.py

KV Enabled

def handler(event, context):

# Read cursor from previous run

last_id = context.kv.get("last_processed_id", default=0)

# Fetch only new items since last run

new_orders = api.get_orders(since_id=last_id)

for order in new_orders:

process_order(order)

# Save progress for the next execution

if new_orders:

new_id = new_orders[-1]["id"]

context.kv.set("last_processed_id", new_id)

print(f"Updated cursor to {new_id}")

return {"processed": len(new_orders)}

zsh • terminal

litelambda-cli v0.1.1

$ pip install litelambda-cli
Successfully installed litelambda-cli-0.1.1

$ litelambda deploy scraper.py --schedule "0 9 * * *" --name "Morning Scraper"
==> Deploying 'Morning Scraper' to LiteLambda Cloud...
✓ Job successfully created!
• ID: c91abfc1-4732-456d-b2d3-2e88f594da81
• Schedule: 0 9 * * * (9:00 AM UTC Daily)

$ litelambda run "Morning Scraper"
✓ Execution triggered. Duration: 842ms • Status: SUCCESS

$ litelambda logs "Morning Scraper"
[09:00:02] Starting background task...
[09:00:03] Fetched 142 items from API. Updated KV cursor.

Official CLI on PyPI • BETA

Prefer your terminal?
Deploy in a single command.

Stay in VS Code, PyCharm, or your local terminal. With litelambda-cli, you can schedule scripts, install pip dependencies, trigger test executions, and stream logs without ever opening a browser.

✓ Pure standard library: Zero heavy wheels or bloated dependencies. Installs in 2 seconds.
✓ CI/CD ready: Schedule and deploy directly from GitHub Actions with a headless token.
✓ Instant log streaming: Watch execution logs directly in your command line.

pip install litelambda-cli

Read CLI Documentation →

Execution Pipeline

From code to logs
in milliseconds.

Every execution follows the same zero-surprise pipeline — sandboxed, logged, and instantly alerted. Nothing runs in the dark.

daily_report.py

import requests, json

def handler(event, ctx):

# Pull last cursor

since = ctx.kv.get("cursor", 0)

data = fetch(since)

ctx.kv.set("cursor", data.id)

Script Ready

Triggered

Sandboxed

Isolated Container

Fresh Docker env · pip cached · CPU + RAM provisioned

Executing... (342ms)

Done ✓

Run Successful

Logged · KV updated · 342ms

execution.log

Pulling image layer... cached ✓

Installing requests==2.31.0... done

Running daily_report.py

> Fetched 58 new records.

KV cursor updated → id:5821

Pristine every run

Fresh container, clean pip environment. No dependency pollution between executions.

Full structured logs

stdout, stderr, return value, duration, and exit code — every run, always searchable.

Instant failure alerts

Slack, Discord, Telegram, or email. Fires the moment a job fails — not when a user complains.

Honest Comparison

"Why not just use a $5 VPS?"

You can. But running periodic Python tasks on an unmanaged server quickly turns into an ongoing maintenance tax.

VPS

The $5 Unmanaged Server

Hours of manual babysitting every month

✕ Security & OS patches: You handle kernel updates, SSH hardening, and CVE patches — or get compromised.
✕ Silent failures: Crons fail silently in /var/log/syslog. You only find out when a user complains.
✕ Disk fill-ups: Unrotated log files and Docker layers slowly eat disk space until the VM locks up at 3 a.m.
✕ Environment drift: System-wide Python upgrades break your dependencies and virtualenvs.

The LiteLambda Way

Set it up in 30 seconds, never patch a box

✓ Zero maintenance: Every execution runs in an ephemeral, pristine container sandbox. Nothing to patch or update.
✓ Instant chat alerts: Native failure notifications straight to Slack, Discord, Telegram, or email.
✓ Built-in KV state: Remember data between runs with context.kv without spinning up Redis.
✓ Structured logging: Searchable logs, status, stdout/stderr, and return values stored automatically.

Built for Developers

Everything you need to automate reliably

Cron Scheduling

Standard 5-field crontab syntax. Run every minute, hourly, or on complex multi-day patterns.

pip Package Support

List your dependencies and we install them. Cached layers make subsequent runs blazing fast.

Full Execution Logs

Every run is logged with stdout, stderr, return value, duration, and timestamp. Debug fast.

Slack Discord Telegram

Alerts where you actually are

Get notified in Slack, Discord, or Telegram the moment a job fails — not just buried in an email inbox. Set it up with a single webhook URL.

Isolated Sandboxes

Each run is isolated in its own container with no shared state. Secure and reproducible by design.

Built-in KV Store Beta

Persist data across runs instantly. Use context.kv.set() and get() without provisioning any external databases.

Manual Runs

Trigger any cron manually to test before going live. No waiting for the next scheduled tick.

Terminal CLI PyPI

Schedule, deploy, test, and stream logs directly from your terminal or CI/CD pipelines with litelambda-cli.

MCP for AI Agents NEW

Let Claude, Cursor, ChatGPT, and Antigravity deploy, schedule, test, and monitor Python jobs autonomously from chat. Explore MCP →

500+

Developers signed up

50000+

Scripts executed

100%

Serverless — zero config

What Developers Say

"I had a working Shopify report script deployed and running in under 5 minutes. I was expecting to spend an afternoon setting up servers and cron jobs."

Thomas B.

E-commerce founder, Berlin, Germany

"I replaced my whole AWS Lambda + EventBridge setup with LiteLambda. No YAML, no IAM roles. The pricing is a fraction of what I was paying AWS."

Marc D.

Backend Developer, Amsterdam, Netherlands

"Finally a cron service that runs actual Python with pip packages. PythonAnywhere's daily limits were killing my automation. LiteLambda just works."

Sarah K.

Solo founder, London, UK

Automate your first task in 60 seconds

Write your Python script, paste it in, and hit deploy. No servers. No DevOps. No surprises.

Get Started — 15-Day Trial

No credit card required

GDPR compliant

Cheaper than AWS Lambda

Cancel anytime

Trusted by developers in Germany, Netherlands, UK, France & more

Accepted Payment Methods

GDPR Compliant

Secured by Razorpay

Connect AI Agents (MCP)

Instant model-to-cron background execution

Configuration copied to clipboard!

Simply paste it into Cursor Settings to complete activation.

Quick 2-step setup:

In Cursor, open Settings > Features > MCP
Click + Add New MCP Server, paste (Cmd+V / Ctrl+V), and Save.

{
  "mcpServers": {
    "litelambda": {
      "url": "https://mcp.litelambda.in/mcp"
    }
  }
}

Launch Cursor App → View full Cursor docs

Option A: Claude Code CLI (Recommended)

Run this single terminal command to register LiteLambda into Claude Code:

claude mcp add litelambda -- https://mcp.litelambda.in/mcp

Option B: Claude.ai Custom Connectors

Endpoint URL to paste into Claude.ai Connectors:

https://mcp.litelambda.in/mcp

Open Claude.ai Connectors →

Add to ~/.gemini/config/mcp_config.json:

{
  "mcpServers": {
    "litelambda": {
      "serverUrl": "https://mcp.litelambda.in/sse"
    }
  }
}

LiteLambda streams MCP events over SSE to all local Antigravity subagents.

OpenAPI / HTTP Action Schema Endpoint:

https://mcp.litelambda.in/mcp

Use this URL when creating Custom GPT actions or AI agent function tools.

Copied to clipboard!

Run your Python in the cloud. On a schedule, on a webhook, or from your AI agent.

How Developers Run Python on LiteLambda

AI Coding Agents

Webhook Backends

State-Aware Monitors

Python Function → API

Custom MCP Tools

Sandbox-as-an-API

The Background Execution Cloud for AI Coding Agents.

AI Coding Agent

LiteLambda Cloud Sandbox

Supported AI Coding Environments

Anthropic Claude

Cursor IDE

Google Antigravity

OpenAI ChatGPT

Ready to empower your AI coding agents?

Scheduled or on-demand.Trigger from anywhere.

Deploy a cron job in under a minute

Write Your Code

Set a Schedule

Go Live

Your job needs to remember things. Now it can.

Prefer your terminal? Deploy in a single command.

From code to logs in milliseconds.

Isolated Container

Pristine every run

Full structured logs

Instant failure alerts

"Why not just use a $5 VPS?"

The $5 Unmanaged Server

The LiteLambda Way

Everything you need to automate reliably

Cron Scheduling

pip Package Support

Full Execution Logs

Alerts where you actually are

Isolated Sandboxes

Built-in KV Store Beta

Manual Runs

Terminal CLI PyPI

MCP for AI Agents NEW

Automate your first task in 60 seconds

Connect AI Agents (MCP)

Option A: Claude Code CLI (Recommended)

Option B: Claude.ai Custom Connectors

Run your Python in the cloud.
On a schedule, on a webhook, or from your AI agent.

The Background Execution Cloud
for AI Coding Agents.

Scheduled or on-demand.
Trigger from anywhere.

Your job needs to remember things.
Now it can.

Prefer your terminal?
Deploy in a single command.

From code to logs
in milliseconds.