Launching today
Devlop Ai

Devlop Ai

AI IDE that writes and flashes STM32 firmware for your board

102 followers

AI coding agents to speed up STM32 embedded development
Devlop Ai gallery image
Devlop Ai gallery image
Devlop Ai gallery image
Devlop Ai gallery image
Free Options
Launch Team / Built With
Framer
Framer
Launch websites with enterprise needs at startup speeds.
Promoted

What do you think? …

Omer
Maker
📌
It will truly bring a breath of fresh air to the embedded development
mostafa kh

@ozkaya anyone who has spent hours digging through stm32 datasheets for pin assignments knows this pain. ai suggesting optimal configurations based on peripheral requirements is a huge time saver.

cubemx import is smart. most embedded projects already have ioc files so migration is easy.

one click compile and flash without external tools is what the stm32 workflow always needed.

two questions: does it support other stm32 series beyond m4 and m7 like f1, f0, g0? and which ai model powers the pin suggestion feature?

Omer

@topfuelauto Great questions,

Yes we absolutely support other series beyond M4 and M7 including F1, F0, and G0. We are also working hard every day to expand our hardware compatibility even further.

On the technical side, we are currently leveraging the Claude infrastructure for our pin suggestion engine. However, we are in the middle of fine tuning our own dedicated model specifically for embedded systems. We’re planning to transition to our proprietary model very soon to provide even more precise and hardware-aware suggestions. Stay tuned

Xiang Lei

How do you handle the "hallucination" problem when dealing with specific datasheet constraints like clock configurations or DMA mappings?

Omer

@lightninglx Great questions,

We don't rely solely on the AI's 'creative' output for hardware-critical tasks like clock trees or DMA mappings. To prevent hallucinations, we use a multi-layered validation approach:

  1. Data-Driven Guardrails: The AI's suggestions are cross-referenced against a structured database parsed directly from official STM32 datasheets.

  2. Rule-Based Engine: We have a validation layer that checks if the suggested pinout or DMA channel is physically available and compliant with the specific MCU's hardware constraints.

  3. CubeMX Integration: By supporting .ioc imports, we ensure that the foundation of the project stays within the 'safe' bounds of official configurations.