Observe → understand
Discover what your space is telling you.
Discover what happens in your space. Log greenhouse temperatures, follow the readings in your IoT dashboard, and see what changes between visits.
Your idea. Connected.
The greenhouse that tells you what changed. The sensor that sparks a product. Bring them to life with a free IoT platform for connected projects.
Warblet turns sensor data into a live IoT dashboard, with routines for email alerts and devices that work together. Connect an ESP32 or another supported board and start exploring what it can do.
Up to 10 devices. No credit card. Nothing is charged today.
Greenhouse sensor
HTTP · MQTT · TCP · UDP
Raw bytes become a reading you can use.
Your routine → a fan command
Compare inside and outside. Decide when to act.
With readings and responses connected, the next experiment is within reach.
Room for your imagination
Observe → understand
Discover what happens in your space. Log greenhouse temperatures, follow the readings in your IoT dashboard, and see what changes between visits.
Notice → respond
Let a reading start something useful. A device can trigger a camera, and a routine can ask for a picture and send an email alert.
Experiment → improve
Bring a prototype's devices into one dashboard. Learn from the readings, change the design, and see what happens next.
From packet to possibility
Get to the part you wanted to build: the experiment, the insight, the behavior. Warblet gives your data a path from the device to the next decision.
Connect your board to the internet, then send a packet over HTTP, MQTT, TCP or UDP.
A short decoder script turns your payload into named fields. Test it against a sample in the browser.
Write routines that use device readings to queue commands, call webhooks or raise alerts.
Build an ESP32 dashboard or connect another supported board. Give your readings names and units, then arrange a view that makes sense for your project.
Chart stored readings to see what changed. Follow incoming messages live and trace how data moves through your project.
Name your devices, check their readings, and send commands from Warblet. Your firmware gives each command its meaning.
Imagine it. Build it. Automate it.
Write a routine that turns incoming data into an email alert or a device command. Build the response around the change you care about.
Build an ESP32 email alert →
Send a camera photo by email →
Want a model in the loop? Routines can call Claude using your own Anthropic API key. You decide where that judgment belongs.
Try asking your connected agent
“Read my sensor’s latest messages. Draft a decoder for this payload, test it on a sample, and show me the fields.”
device_messages → decoder_test → decoder_put
An example request, using real Warblet MCP tools.
AI agents + MCP for IoT
Use MCP to let your agent inspect device messages, write and test decoders, create routines, and queue commands. Spend the conversation on what you want to accomplish.
Connect supported chat clients through sign-in and consent, without a Warblet API key. Claude Code and other MCP clients can connect with an API key.
Connect your agent →Get something real talking
Try a ready-made demo on a supported board with the browser flasher. Review the firmware and setup before you write, then watch for its first packet on your account.
Most demos use Wi-Fi; the Thread demo needs a border router with internet access. Use another link, such as Ethernet, if it can carry HTTP, MQTT, TCP or UDP to Warblet.
Explore source examples in MicroPython, ESP-IDF/C, STM32 bare-metal C, an Arduino sketch and portable C. Adapt a demo to your sensor and get the next experiment moving.
Plug in your board over USB for setup and choose its port. It also needs an internet connection to send data.
Choose a compatible demo. Check its requirements, source and settings before pressing GO.
Use the demo as a starting point. Follow the incoming data, then try the next sensor, command or routine.
Demo catalog chips: ESP32, ESP32-S3, ESP32-C3, ESP32-C5, ESP32-C6, STM32WBA65 and STM32F722.
USB setup needs a compatible desktop browser, such as Chrome or Edge. Some demos need an existing runtime, a relay or a Thread border router; the setup card explains what to prepare.
Your board. Your language. Your stack.
Start with an ESP32 demo or connect another supported board. Bring your own firmware, use C or ESP-IDF, or explore MicroPython. Send readings over a supported connection and build the behavior that makes the project yours.
Choose for the real world
Choose authentication per device, then choose whether to require encryption in transit. Match the settings to the data, the network and what your device controls.
Accept packets without device authentication. Anyone who knows the device ID can send data as that device.
Read the Open threat model →Require a device secret. Protect it in storage and in transit: someone who captures the token can use it.
Read the Token threat model →Verify payloads with HMAC-SHA256. Protect the signing key. Signing does not encrypt data, and replay protection requires a nonce.
Read the Signed threat model →Enforce TLS rejects plaintext connections for that device. Use HTTPS or MQTT over TLS. Encryption protects data in transit; it does not replace device authentication.
How Enforce TLS works →Pricing
Start free with the flasher, cross-device routines, decoders, downlink, all four transports and every security level included. If you go over a limit, we throttle you and tell you. We do not quietly bill you.
$0up to 10 devices
The whole platform, ten devices, no credit card. Start with a sensor, a prototype fleet, or the first version of a product.
per device / month
For the project that grew past ten boards. You pay for what is actually plugged in, and nothing else changes.
volume, per device
Shipping units with your name on them. Claim codes by the batch, provisioning for a production line, and a number we size with you.
Creator and Business are invite only, and nothing is charged today. Start on the free tier — when you move up, your devices, decoders and routines carry straight over.
Build on every discovery
Start with a personal project. Follow it toward a prototype, a startup, a team. See what you can build when your devices have somewhere to talk.