Spring 2026 Updates: Faster AI, Better Layouts, Smarter Sourcing

You research parts across half a dozen tabs. You build a plan in your head. You lay out the same classes of routing over and over. You wait on tools. Then, just when the design is taking shape, sourcing changes underneath you.

This spring, we’ve made Flux faster, more capable, and more practical across the full hardware design workflow.

Today, we’re introducing the Spring 2026 Update: a major upgrade to how Flux helps you research parts, build schematics, lay out boards, and design with sourcing in mind.

With these updates, we bring together a faster, more self-correcting AI agent, major improvements to AI Auto-Layout, intern-like multi-step execution, and real-time supply chain visibility built directly into the design workflow. The result is a faster path from idea to board, with less manual busywork and fewer late-stage surprises.

From prompts to real execution

Flux now helps with more of the hardware design workflow end to end, from planning and part research to schematic work, layout, and sourcing-aware decisions.

You can describe what you want to build in terms of goals, constraints, interfaces, power targets, and preferred components, and Flux can turn that into a step-by-step plan spanning parts research, schematic creation, layout, checks, and review milestones. As it works, it shares progress, explains decisions, and checks in at key moments so you can steer the design without micromanaging every step.

That shift matters because hardware work is rarely a single action. Real progress comes from coordinating many decisions across planning, implementation, verification, and iteration. The Spring Update is about making that full workflow feel dramatically more natural inside Flux.

A much faster, more self-correcting agent

At the core of this update is a major upgrade to how Flux executes design work.

The agent was rebuilt around a new architecture, execution model, and verification system designed to improve three things at once: speed, correctness, and reliability.

One of the biggest changes is that the agent now checks its own work continuously while it operates. Instead of generating a result and leaving validation until the end, it runs ERC and DRC checks as it goes, catching issues like floating pins, missing connections, and rule violations in real time. That means cleaner schematics on the first pass, fewer manual fixes, and fewer wasted iterations.

It’s also more adaptive during execution. Rather than following a fixed plan from start to finish, the agent evaluates progress after each step and adjusts as it works. If it encounters a conflict or finds a better path forward, it updates its approach and continues. In practice, that makes it more effective across both lightweight tasks and more complex multi-step jobs.

Better layouts with less cleanup

We’ve also continued to improve AI Auto-Layout based on real usage and feedback.

This update focused on making results cleaner, easier to review, and more manufacturable. Routing now converges faster, with shorter traces, fewer vias, and paths that better follow placement intent. Layouts are more legible, and the system more reliably produces results that need less cleanup before handoff.

For many teams, that is exactly the right division of labor: human judgment on what matters most, AI speed on the busywork in between.

Supply-chain-aware design, built in

Good hardware design depends not just on performance, but on choosing parts that are actually available.

Most EDA tools still treat sourcing as something you check after the design is already locked. But by then, a BOM change can ripple through the whole project. Flux’s approach is different: pricing and availability stay connected to the project while you work, so supply constraints surface early, when design decisions are still flexible.

With the Spring Updates, that means you can:

• See pricing and availability directly in part search
• Ask Flux AI to find in-stock parts based on your real design requirements
• Ask for replacements that fit your context
• Inspect real-time distributor stock for parts already in your design

This is a meaningful change in workflow. Instead of designing first and negotiating with reality later, you can make sourcing-aware decisions as the design takes shape.

Starting From Proven Foundations

Every hardware engineer has rebuilt the same circuits dozens of times. Voltage regulators, programming headers, decoupling networks, reference designs for common microcontrollers.

Flux now makes it possible to start from reusable design foundations instead. Engineers can fork reference designs, microcontroller platforms, and manufacturer rule templates, then adapt them for their specific application. Teams can also build internal templates that encode their own design standards, compressing the gap between a new idea and a working architecture.

This is the last of the five updates, and the most foundational in terms of workflow change, not because the technology is the most complex, but because it reframes where design work actually begins.

Try the Spring Updates

Together, these improvements make Flux more useful at every stage of hardware design.

If you’ve used Flux before, now is a good time to run a workflow you already know well and feel the difference for yourself.

Try it in Flux today!