Companies·2 min read·TNW

Sereact Raises $110M Series B to Scale 'Robot Brain' VLAMs Across BMW, Daimler, and US Logistics

German robotics startup Sereact closed a $110M Series B led by Headline to scale its vision-language-action models — already running in BMW, Daimler Truck, and Mercedes plants — and to push into the US warehouse market.

Sereact Raises $110M Series B to Scale 'Robot Brain' VLAMs Across BMW, Daimler, and US Logistics
Share:

Stuttgart-based Sereact has closed a $110 million Series B led by Headline, with new participation from Bullhound Capital, Felix Capital, and Daphni alongside existing backers. The round, announced April 27, vaults the four-year-old company into the upper tier of European robotics startups and gives it the balance sheet to launch a long-telegraphed expansion into the United States warehouse market.

Sereact builds Vision-Language-Action Models (VLAMs) — the robotics analogue of the multimodal LLMs that have absorbed most of the AI capital cycle. Where a generative model maps a prompt to a string of tokens, a VLAM maps a camera feed and a natural-language instruction to a physical action plan. Sereact's twist is a predictive layer that lets a robot evaluate whether a planned grip will damage the object before its gripper closes, the kind of soft-failure intuition that has historically separated human pickers from automation.

The customer roster is the company's most credible asset. BMW Group, Daimler Truck, Mercedes-Benz, Bol, MS Direct, and Active Ants are running Sereact systems in production, and the company says more than 200 deployments have completed over a billion picks with human intervention required in roughly one of every 53,000 — a reliability figure that, if it survives independent audit, would put Sereact among the most production-hardened embodied-AI platforms in the world.

Founders Ralf Gulde (CEO) and Marc Tuscher (CTO), both former University of Stuttgart AI researchers, have walked a deliberately European funding path: a $5 million seed in 2023, a Creandum-led €25 million Series A in January 2025, and now $110 million from a syndicate that mixes Berlin, Paris, and San Francisco capital. Headline's lead role and Bullhound's entry signal an explicit US push — Sereact has flagged American logistics as the next frontier for its Cortex generation of models, which add a world-model simulation layer letting robots try multiple candidate actions in silicon before committing to one in steel.

The timing is sharp. Embodied AI funding hit a record pace in Q1 2026 as Sereact's domestic rival Figure scaled, and US warehouse operators reacted to a tight labor market by accelerating automation budgets. With a billion-pick production track record and a frontier model architecture that maps cleanly onto humanoid platforms, Sereact arrives in the American market with a story that Headline clearly believes is fundable several rounds further. The Series B is for scaling Cortex and shipping it through US distribution — the Series C, if the deployment curve holds, will be about humanoids.

Want AI news before everyone else?

The morning's most important AI stories, straight to your inbox. No fluff.

Related Articles

OPENAI'S FIRST CUSTOM CHIP “Jalapeño,” an LLM-inference ASIC built with Broadcom LLM JALAPEÑO Inference, not training Designed in ~9 months Co-built with Broadcom Better performance-per-watt Ships from late 2026 BITSMINDS.COM
Companies

OpenAI Unveils Jalapeño, Its First Custom AI Chip

A NOBEL LAUREATE JOINS ANTHROPIC AlphaFold co-creator John Jumper leaves Google DeepMind Google DeepMind Anthropic JJ Nobel Prize · Chemistry 2024 AlphaFold: 200M+ structures Anthropic’s AI-for-science push BITSMINDS.COM
Companies

Nobel Laureate John Jumper Leaves DeepMind for Anthropic

ENGRAM LAUNCHES WITH $98M A "learned-memory" layer that runs enterprise AI on up to 100× fewer tokens FRONTIER CONTEXT 100% of tokens ENGRAM MEMORY ~1–10% of tokens up to 100× fewer tokens Backers: General Catalyst · Kleiner Perkins · Sequoia Customers: Microsoft · Notion · Harvey BITSMINDS.COM
Companies

Engram Launches With $98M for 'Learned Memory' AI