Reflection AI funding, valuation and investors

Overview

Reflection AI is a New York artificial-intelligence company founded by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou. It develops open-weight foundation models, coding and agent systems, and infrastructure for organizations and governments seeking greater control over model deployment.

Why Reflection AI is interesting

Reflection sits at the intersection of open-weight AI, sovereign computing and Nvidia’s expanding investment ecosystem. Reported acquisition talks test whether a strategically important model supplier remains independent or becomes more tightly integrated with the dominant AI-chip platform.

Product & use cases

Reflection’s flagship Beam family is designed for coding, reasoning and agent workflows. The company says Beam uses a 501-billion-parameter mixture-of-experts architecture with 23 billion parameters active per token, and plans to make model weights and supporting materials available after safety testing. Its broader “AI factory” positioning targets institutions that want localized, customizable model ecosystems.

• Enterprise coding, software-development and agent workflows
• Localized or sovereign AI deployments for governments and strategic industries
• Custom model fine-tuning and controlled on-premises or private-cloud deployment
• Research and commercial development using downloadable model weights

Key facts

• Founded in 2024 by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou.
• Raised $2 billion in October 2025 at an $8 billion valuation, led by Nvidia.
• The Financial Times reported that Reflection was last valued at $25 billion in a March 2026 financing.
• Introduced the 501-billion-parameter Beam open-weight model on October 5, 2026.
• Nvidia is reportedly considering an acquisition, acqui-hire/licensing structure, additional equity investment or deeper compute partnership; no deal has been agreed.

Funding history (newest first)

Investors in our directory

Funds linked from Reflection AI's profile — open a fund page for stage focus and related deal articles.

Competitive landscape

Edge: A large capital base, deep Nvidia relationship, experienced DeepMind founders and a strategically differentiated open-weight model strategy give Reflection access to compute, distribution and government-facing partnerships that smaller open-model labs may struggle to match.

Reflection competes with U.S. and European open-model developers such as Meta’s Llama ecosystem and Mistral AI, and with Chinese open-weight leaders including DeepSeek, Z.ai and Alibaba’s Qwen. It also overlaps with closed frontier labs when enterprise and government buyers compare capability, deployment control, safety and total compute cost. Nvidia’s role as investor, supplier and possible acquirer creates distribution advantages but also concentration and governance risk.

• Meta Llama direct

  Large U.S. open-weight model ecosystem with broad developer distribution.

• Mistral AI direct

  European frontier-model company offering open and commercial model families.

• DeepSeek direct

  Chinese open-weight model developer central to Reflection’s strategic positioning.

• Z.ai direct

  Chinese model developer whose GLM-5.2 is a stated Beam comparison point.

• Alibaba Qwen direct

  Widely adopted Chinese open-weight model family.

• OpenAI adjacent

  Closed frontier-model provider competing for enterprise and government workloads.

Notable stories

• Reflection began with autonomous coding systems before expanding its pitch toward frontier open intelligence and institution-specific AI factories.
• The company’s Nvidia relationship spans capital, compute and go-to-market support, making any transaction strategically meaningful beyond the purchase price.

Industries

Market / IPO context

Related funding articles

Venture Capital Tracker pieces that cover Reflection AI's financing or category context.

• Nvidia Explores Reflection AI Deal After Beam’s Open-Weight Debut