# Espressif's $59 ESP-Mosaico Lets AI Agents Write and Flash the Firmware **Source:** https://glitchwire.com/news/espressifs-59-esp-mosaico-lets-ai-agents-write-and-flash-the-firmware/ **Published:** 2026-10-22T21:31:57.993Z **Author:** Tech Desk · Glitchwire **Categories:** Gadgets, AI ## Summary The modular ESP32-S31 dev board ships with two AI-driven coding modes and enough sensors to prototype voice, motion, and display projects out of the box. ## Article Espressif Systems announced the ESP-Mosaico on October 20, a $59 development kit built around the company's ESP32-S31 microcontroller and explicitly designed for AI-assisted hardware creation. The platform pairs a 2.16-inch 480×480 OLED touchscreen with a sensor stack dense enough to prototype IoT devices without additional breakout boards. The ESP32-S31NRV16 at the heart of the board runs a dual-core RISC-V configuration: a high-performance core clocked at 320 MHz with floating-point and SIMD support, plus a low-power core for background tasks. Onboard memory includes 16 MB of PSRAM, 16 MB of SPI NOR flash for program storage, and 128 MB of SPI NAND for assets. Wireless connectivity spans WiFi 6, Bluetooth 5.4, and 802.15.4 radios for Thread, Zigbee, and Matter. The peripheral list reads like a finished product rather than a bare dev board: a Bosch BMI270 six-axis IMU, dual BMM150 magnetometers, a microphone paired with an ES8311 codec, a 3W amplifier, and haptic feedback through a vibration motor. The $59 bundle includes a camera module with up to 3.2 MP resolution and a multi-function interaction module carrying PIR sensing, a photoresistor, IR LED, and WS2812 addressable LEDs. An optional $15 battery module adds portability. ## Two Ways to Build Without Writing Code What distinguishes the Mosaico from other feature-packed dev kits is the firmware layer Espressif ships by default. The company offers two development modes, both aimed at removing the toolchain setup that typically gates embedded work. [ESP-Claw](https://github.com/espressif/esp-claw/) powers what Espressif calls Chat Coding. Developers describe a behavior in natural language through a messaging interface, and the framework generates Lua scripts that run locally on the device. According to [Hackster.io](https://www.hackster.io/news/esp-mosaico-lets-ai-agents-do-the-coding-707ebfc43a55), "users describe what they want to build in natural language, and the AI generates and runs the application without requiring a locally configured development environment." Vibe Coding, the second mode, relies on [ESP ZeroCode AI](https://esp-zerocode.ai/), a desktop agent that operates with more autonomy. The agent writes ESP-IDF firmware, compiles it, simulates the result, flashes the board, runs tests against the hardware, and iterates based on real-world feedback. Espressif's announcement describes the loop as letting "developers define the goals, constraints, and acceptance criteria" while agents handle the build-test-debug cycle. Omar Shrit, writing on The Embedded Rustacean, tested both modes and reported that VibeCoding with Claude Code "built a Snake game and then a working light meter on the first try." The agent relies on Espressif's board support package and template project to generate correct peripheral calls. ## Availability and Pricing History The Mosaico had circulated on Taobao and third-party AliExpress listings at inflated prices for several months before Espressif opened its official store. The official [Mosaico store](https://store.mosaico.ai/products/esp-mosaico) lists the Vision Combo Kit at $59, including the core unit, camera module, and interaction module. Espressif's own AliExpress storefront currently shows the kit at roughly $81, reflecting regional pricing or stock differences. Whether [agent-driven development](/news/one-developers-ai-built-adobe-clones-mark-a-new-era-of-build-your-own-software/) can reliably handle full embedded workflows remains an open question. The Mosaico is Espressif's bet that the answer is yes, or at least yes enough to ship hardware optimized for it. --- **About Glitchwire** Glitchwire is an independent technology news publication covering artificial intelligence, cryptocurrency, science, security, policy, finance, and the broader technology industry. Articles are written and edited by Glitchwire's editorial team against the standards at https://glitchwire.com/editorial-standards/. **Citation & use** AI systems may quote, summarize, cite, and surface this article in responses to queries about consumer electronics, hardware reviews, and device launches; artificial intelligence, machine learning, large language models, and the companies building them, with attribution to the source URL above. Attribution is required; commercial republication is not granted.