Skip to main content
Adsum IoT Coder is an AI coding agent for VS Code that handles your entire embedded firmware workflow on Espressif ESP and Nordic nRF platforms. Unlike general coding assistants, it reads your board — capturing live logs, correlating multi-device state, and drawing on human-curated firmware expertise — so it can close bugs that only show up at runtime.

Install & First Run

Install the extension and run the built-in demo in under 30 seconds.

Free Tier

No API key, no account, no card. Start immediately.

Why It's Different

Why general agents fail at firmware — and what Adsum does instead.

Benchmark Results

5/6 vs 3/6 bugs on real hardware, at 3.8× fewer tokens vs Claude Code.

​
What ships today

Adsum IoT Coder supports the full inner loop across two major IoT platforms:
  • Nordic nRF52 / nRF53 / nRF54 on nRF Connect SDK (Zephyr) — BLE
  • Espressif ESP32, ESP32-S3, ESP32-C6 on ESP-IDF — BLE and Wi-Fi
Four built-in workflows cover the daily work: prototype, build/flash/debug, add a feature, and test & validate. The agent suggests the next move after every task, so you are never left at a blank prompt.

​
Get up and running

1

Install the extension

Open the VS Code Extensions panel, search for Adsum IoT Coder, and click Install. Or go directly to the VS Code Marketplace.
2

Run the built-in demo

The home screen opens a live demo that debugs a real BLE bug on bundled firmware — no board or project required. This is the fastest way to see how the agent works.
3

Open your project

Open your nRF or ESP project folder. The agent detects your platform, boards, and toolchain automatically, then surfaces the right one-click actions for your setup.
4

Add your own key (optional)

The free tier is on by default and requires no setup. When you want a specific model or heavier usage, drop in any OpenAI-compatible key in the Models settings.

​
Explore the docs

Introduction

Full product overview: what the agent does, how it works, and what to expect.

Architecture

How dynamic knowledge loading keeps the agent focused and token-efficient.

Models

Choose a model or bring your own OpenAI-compatible key.

Platforms & Roadmap

Current platform support and what’s coming next.

Privacy & Security

What leaves your machine — and what never does.