Multiverse Computing raises $570M Series C funding at $1.7B valuation for AI model compression technology
Funding ● Confirmed 92% confidence first seen
Multiverse Computing, a Spanish startup specializing in AI model compression, closed a $570 million Series C funding round at a $1.7 billion valuation, co-led by Forgepoint Capital, Bullhound Capital, and BNP Paribas's Solar Impulse Venture Fund. The company's CompactifAI technology uses tensor network techniques to reduce large language model sizes by 50-95% while maintaining accuracy, enabling efficient inference on edge devices like smartphones and laptops. The funding will support expansion of the company's model library, R&D acceleration, and geographic expansion into Asia, the Middle East, and North America.
The deal
Multiverse Computing $570M Series C · announced 27 Jul 2026
Investors Forgepoint Capital Bullhound Capital BNP Paribas Solar Impulse Venture Fund European Innovation Council
Deal terms as reported in the coverage below.
Decision brief
- What changed
- Multiverse Computing, a Spanish startup, closed a $570 million Series C funding round at a $1.7 billion valuation to scale its CompactifAI model-compression technology, which uses tensor network techniques to shrink large language models by 50-95% for edge deployment.
- Why it matters
- This signals growing investor conviction that AI inference is shifting from centralized cloud/data-center dependence toward edge devices (phones, laptops, drones), which could reshape infrastructure cost structures and vendor strategies for enterprises deploying AI at scale. A ~5x valuation jump from the prior round also indicates intensifying competition and capital concentration in the model-efficiency/compression segment, a category CTOs and CEOs should track when planning AI deployment architecture and vendor partnerships.
- Evidence
- Three independent outlets (Sifted, Tech.eu, SiliconANGLE) consistently report the $570M raise and $1.7B valuation with the same lead investors (Forgepoint Capital, Bullhound Capital, BNP Paribas SIVF), though Sifted's initial reporting cited a differing €500m/€2bn figure, suggesting the deal terms evolved or were reported at different stages.
- What remains uncertain
- The compression accuracy claims (50-95% size reduction with 'minimal accuracy loss') are sourced from company statements relayed by press, not independently benchmarked; the 5x valuation increase from the June 2025 round is asserted by one outlet without clear methodology, and real-world enterprise adoption or customer traction is not detailed in the coverage.
- Monitor next
- Watch for independent third-party benchmarks or enterprise customer deployments of CompactifAI-compressed models (e.g., Llama 3.3 70B on consumer hardware) that validate the compression claims outside company-provided data.
Analytical support, not advice — assumptions and open questions stated above.