Eclipse Ventures Raises $1.3 B to Build a Portfolio of Physical‑AI Startups

Why It Matters

The raise underscores a shift in venture capital toward capital‑intensive, hardware‑focused AI, a segment that has lagged behind software‑only AI despite its potential to transform manufacturing, logistics and defense. By committing $1.3 billion to a build‑co model, Eclipse is betting that operational expertise can de‑risk the long development timelines typical of physical AI, offering limited partners a new avenue for diversification. If Eclipse can demonstrate that its hybrid investor‑builder approach yields commercially viable products, it could catalyze a wave of similar funds, prompting a reallocation of capital from pure‑software AI startups to those that embed intelligence in the physical world. This would accelerate the adoption of autonomous robots, smart factories and AI‑enhanced infrastructure, reshaping industrial productivity on a global scale.

Key Takeaways

• Eclipse Ventures closed a $1.3 billion raise on April 7, 2026.
• Fund VI: $720 million for early‑stage physical‑AI startups; Early Growth Fund III: $591 million for scaling companies.
• Portfolio includes Cerebras, Arc, Redwood Materials, Wayve and Bedrock Robotics.
• The raise is the largest in Eclipse’s 11‑year history, exceeding the $1.23 billion round in 2023.
• Eclipse’s build‑co model targets sectors such as robotics, defense, energy and advanced manufacturing.

Pulse Analysis

Eclipse’s latest fundraise arrives at a moment when the AI market is saturated with software‑only bets, many of which are chasing short‑term revenue streams. Physical AI, by contrast, promises deeper, structural impact on industries that have historically lagged in digital transformation. The firm’s decision to split capital between early‑stage discovery and near‑term growth reflects a nuanced understanding of the capital needs across the hardware development lifecycle. Early‑stage funding can address prototype and proof‑of‑concept phases, while the growth vehicle can bridge the gap to commercial deployment, a stage where many hardware startups traditionally stumble.

Historically, venture capital has struggled to fund hardware ventures because of longer ROI horizons and the need for specialized engineering talent. Eclipse’s build‑co approach mitigates these challenges by internalizing the talent‑sourcing and product‑definition processes, effectively reducing the time to market. If the firm can replicate the success of past portfolio companies—Cerebras’s rapid scaling of AI chips, for example—it could validate a new template for venture capital in deep‑tech domains.

However, the model is not without risk. Physical AI projects demand large capital expenditures, complex supply‑chain coordination and often navigate stringent regulatory environments, especially in defense and energy. Eclipse’s ability to secure strategic partnerships with industrial incumbents will be a decisive factor. Moreover, the mention of Algeria’s nascent deep‑tech ecosystem highlights a broader geographic challenge: scaling such a model globally will require ecosystems that can support advanced manufacturing and talent pipelines. Success in the United States and Europe may not automatically translate elsewhere, limiting the firm’s global impact unless ecosystem development keeps pace.

Overall, Eclipse’s $1.3 billion raise could be a bellwether for a new wave of venture capital that embraces the capital intensity of hardware AI. The firm’s performance over the next 12‑18 months will likely shape investor sentiment toward similar funds and could accelerate the convergence of AI and the physical world.