Walden Robotics launches with $300M — and robots that have been on a Toyota line since February

Walden Robotics came out of stealth on July 15, 2026 with $300 million in seed funding at a $1.1 billion valuation, co-led by Toyota — participating through Toyota Motor Corporation, Toyota Invention Partners and Toyota Ventures — and Deviation Capital. NVIDIA, Boeing, AE Ventures, Samsung Ventures, Prologis Ventures and CoreWeave Ventures joined, along with Calibrate Ventures, Colle Capital, Shine Capital, NextView Ventures, Squarepoint Capital, One Madison Group, KAS Venture Partners and Menlo Ventures.

The Cambridge, Massachusetts company was founded in January 2026 as a spinout from Toyota Research Institute. Co-founder and CEO Russ Tedrake is an MIT professor and TRI's former Senior Vice President of Large Behavior Models. Since February, Walden says, its general-purpose robots have been doing useful work in production at a Toyota plant in North America — from first pilot to real work in under two months, per the company.

What is actually being underwritten

Most nine- and ten-figure physical-AI rounds in 2026 have priced a research team and a roadmap. This one prices a deployment that already exists, on a line owned by the lead investor. That is a narrower claim than it sounds — a captive first customer is the easiest customer — but it inverts the usual sequence, where a company raises, builds, pilots, and then spends two years failing to convert pilots into contracts.

Providing real value to customers and building a robust and scalable business requires a deep understanding and respect for how manufacturing is done today. — Russ Tedrake, co-founder and CEO, Walden Robotics

The technical lineage

Walden's stack rests on work the founding team helped originate at TRI: Diffusion Policy and Large Behavior Models (LBMs). The pitch is that LBM-driven robots pick up new tasks quickly and improve through real-world practice rather than reprogramming — which is the same wager Google DeepMind is making with Gemini Robotics 2 and that China's data factories are trying to brute-force with demonstration hours.

The difference is where the data comes from. Walden's model improves on a working line, in front of the people whose jobs it touches, rather than in a teleoperation warehouse. If that loop closes, the deployment is the data strategy.

The investor list is the strategy

Read the cap table as a target market. Toyota is automotive. Boeing is aerospace. Samsung is semiconductors and electronics. Prologis is logistics real estate. NVIDIA and CoreWeave are compute. Walden names automotive, aerospace, semiconductors, electronics, logistics and life sciences as strategic-partner industries — each of the first four has a corresponding investor. That is a company buying distribution with equity, in sectors where a robotics vendor's hardest problem is getting through the door of a plant at all.

Toyota CTO Hiroki Nakajima framed the investment around kaizen and jidoka and "keeping people at the center" — language that matters more than it reads, because the Toyota Production System's stance on automation has historically been that machines assist skilled workers rather than replace them. It is also a hedge: Toyota gets an option on general-purpose robotics without committing its own manufacturing roadmap to a specific machine.

Why it matters

2026's physical-AI capital has flowed overwhelmingly toward humanoid form factors and toward China, where rounds have been both larger and more frequent. Walden is a US company that raised at unicorn pricing on the strength of an operating deployment rather than a demo reel, and it did so without leading with a humanoid.

The number to watch is not the next raise. It is whether a second, non-Toyota customer runs Walden robots in production, and whether the company discloses task-level performance rather than deployment anecdotes. Until then the $1.1 billion is priced on one factory and a research pedigree.