The AI hardware engineer. One agent with the right models, tools and parts for each step. Start from an idea or your own files and build working hardware.
Today, we’re excited to share that Flux has raised $37M in new funding, including a $27M Series B led by @8VC, with participation from @BainCapVC, @Liquid2V, and @OutsidersFund.
To the Flux community: we couldn’t have done this without you! Your projects and feedback shape what we build.
Draw the change you want. Flux plans it.
We circled a battery on a pair of AI glasses in Flux's 3D view and asked for one battery, Tag-Connect programming and charging contacts. Flux came back with a four-step plan.
Markup is live: flux.ai/p/blog/markup-mode
Our first flex boards just landed.
Designed in Flux and built to bend, with “Made with Flux” on every one.
Flex and rigid-flex are now in Flux: flux.ai/p/blog/flex-and-rigi…
2026 is the year AI learned to engineer the physical world.
Within a year, I believe AI will design robots, satellites, and vehicles.
After 7 years building Flux, here's why I think hardware engineering and the CAD industry are about to change forever.
We ran Fable 5.2 on real hardware overnight.
What stood out: it places parts like an EE would. Room for ESD around USB-C. CC resistors right where they belong.
Now live in Flux as a Preview model. Pick it yourself, or let Flux choose the right model for each job.
As someone who spends six figures a month benchmarking AI models for hardware engineering, I can tell you: there is no single best model!
Every model has strengths and weaknesses. The best model for one task can be terrible at another.
This is why harnesses like Devin for coding and Flux for hardware are so powerful. They orchestrate multiple models, routing each task to whichever model is best suited for it.
If you’re exclusively using Claude Code, Codex, or any single model, you’re not experiencing the cutting edge of AI.
I think domain-specific harnesses are where the real breakthroughs will happen. Systems that deeply understand the problems they’re solving and know how to get the most out of whatever models and tools are available.
Catch parts with no datasheet before fab.
Flux review now lists every part in your design that needs a datasheet but does not have one.
IoT builders can check that each sensor and radio on a smart home board has its datasheet before they order.
Orbit the 3D view around any point.
Press and drag in the Flux 3D view, and the camera now orbits around the spot under your cursor.
Drone builders can circle one motor connector on a flight controller without the board swinging out of view.
Zoom and pan any diagram in fullscreen.
Open an image, SVG, chart or diagram from the Flux chat in fullscreen, and you can now zoom in and drag it around.
Any builder can read the small labels on a block diagram or pinout without leaving the chat.
Editing a part is the hardest first job in CAD. Now you just ask.
Bring a STEP, IGES or BREP file into Flux and say what to change. Flux edits it, keeps the holes in place, and saves a new STEP file.
Taller power supply? Ask for a deeper flange with vent slots.
Bring your IGES and BREP models into Flux.
Flux now accepts IGES and BREP files as 3D models, the same way it takes STEP.
Robotics builders can upload a gripper or chassis part from their CAD tool without converting it first.
Design an entire multi-board device in one Flux project.
Flux plans the architecture, designs PCBs in parallel, then assembles them in 3D to check fit between boards and inside the enclosure. All before routing a trace.
Think drones with separate flight controllers, ESCs, and radios.
One project. Multiple boards. One AI hardware engineer.
Tomorrow night we're hosting the 138th SF Hardware Meetup at Flux HQ, part of SF Tech Week.
Talks, Q&A and an open mic on supply chain and the future of manufacturing.
Thu, Oct 8 · 6:30–9:30 PM