StopDown
Analysis · 9 March 2026OpenClawOpen-source agentic automation framework

OpenClaw users build AI agents for real-world tasks

By · Editor ·

  • OpenClaw is an open-source framework that lets AI agents interact with browsers and apps like a human
  • Founders and engineers built personal agents for dental labs, budget tracking, and enterprise simulation
  • Costs range from $6/month for basic setups to over $200/month for heavy automation with premium models

What happened

  • OpenClaw is an open-source agentic automation framework that was formerly called Clawdbot and Moltbot.
  • It was acquired by OpenAI, and founder Peter Steinberger joined OpenAI.
  • The framework lets AI agents interact with browsers, dashboards, and apps like a human using screenshots and dynamic navigation.
  • Users can install modular 'skills' to expand capabilities without complex workflow engineering.
  • Raman Choudhary, a Jaipur-based engineer, built an OpenClaw agent to automate dental lab workflows — portal logins, file downloads, and marketing content updates via WhatsApp.
  • Aditya Pratap, a fourth-year BTech student, built a multi-agent orchestration system that routes tasks across eight specialised agents (sales outreach, GitHub monitoring, database queries, daily briefings).
  • Soham Mullick, a lead data scientist at Teradata, built an OpenClaw budget tracker for Union Budget 2026 that pushed updates every 15 minutes tailored to Karnataka and salaried-class impact.
  • Harshil Mathur, Razorpay cofounder, built a mini 'Jarvis' using OpenClaw that reads sleep/recovery data, generates morning briefings, orders food on Swiggy, auto-sends gifts, and turns off lights when he is in bed.

Why it matters

  • Meta's chief AI officer Alexandr Wang said personal agents are AI's next opportunity.
  • OpenClaw's appeal is its autonomy — it interacts with browsers, dashboards, and apps like a human.
  • The framework is raw infrastructure: powerful but flaky, with costs that scale unpredictably and significant security risks.
  • Harshil Mathur described OpenClaw as 'like an untrained intern, extremely intelligent, but inexperienced'.

The business

  • OpenClaw is described as 'AI that actually does things' — it manages calendars and builds Reddit-style social networks of AI agents.
  • It is open source and free; costs come from infrastructure and AI usage.
  • Basic personal setups cost $6–$13 per month, small business workflows $25–$50 per month with mixed model routing.
  • Heavy automation with premium models can exceed $200 per month.
  • The framework supports a skills ecosystem for modular extensions.
  • It runs in sandboxed environments to avoid exposing company data.

The founders

Raman Choudhary

Founder of Dent Node
  • Jaipur-based engineer, ex-senior software engineer in Bengaluru.
  • Left his job to join his dentist father's venture.
  • Founded Dent Node in November 2025.

Aditya Pratap

Fourth-year BTech student
  • Built a multi-agent orchestration system with OpenClaw.

Soham Mullick

Lead data scientist at Teradata
  • Bengaluru-based.
  • Built OpenClaw budget tracker for Union Budget 2026.

Harshil Mathur

Cofounder of Razorpay
  • Built a mini 'Jarvis' using OpenClaw and MCPs.

Manu Mishra

Senior engineer at AWS
  • US-based.
  • Simulating an enterprise with a single human as CEO and OpenClaw agents as virtual employees.

Akash Avvaru

Engineer
  • Hyderabad-based.
  • Runs AI 'employee' agents for freelance web development and consulting.

In numbers

nearly ₹1,000API credits spent in a single day · 2026
$100–$200Daily cost using Opus model · 2026
$30–$50Daily cost after model routing · 2026
~70%Cost reduction from model routing · 2026
$6–$13Basic personal setup monthly cost · 2026
$25–$50Small business workflow monthly cost · 2026
exceeds $200Heavy automation monthly cost · 2026
2,700+Emails sent without spam · 2026

Words to know

agentic automation
AI systems that can autonomously perform tasks by interacting with software and websites, like a human would.
MCP (Model Context Protocol)
A protocol that allows AI models to access external tools and data sources securely.
sandboxed environment
A secure, isolated space where code runs without affecting the rest of the system.
model routing
Using cheaper AI models for simple tasks and more expensive ones for complex reasoning, to reduce costs.

The numbers around it

AI & Deep Tech in March 2026: 11 rounds with a disclosed amount, $79.8 million in all; the biggest was Deccan AI ($25 Mn). Every sector this year →

Read it as a card in the app

All startup news on 9 March 2026 →

Is this OpenClaw? Get your “Featured on StopDown” badge →