Reflection launches Beam, an open-weight AI model aiming to match Chinese rivals with lower compute costs

Reflection AI, a Brooklyn-based startup founded in 2024, has officially launched Beam, its first frontier open-weight model. The company claims Beam matches leading Chinese open models on advanced reasoning benchmarks while consuming far less compute, a pitch that positions it as a potential Western counterweight to DeepSeek, Qwen, and Z.ai. The announcement confirms Axios reporting from the weekend that a launch was imminent.

What Beam Actually Is

Beam is a text-only mixture-of-experts model with 501 billion total parameters and 23 billion active parameters. It was pretrained on 23.8 trillion tokens and supports a 1 million token context window. For comparison, Z.ai's GLM-5.2 carries roughly 744 billion total parameters with 40 billion active.

Reflection says Beam was trained on high-compute reinforcement learning to handle reasoning, coding, and agentic tasks at "a fraction of the token cost and inference time compute" of rivals. The company describes it as a "workhorse model" for enterprises, the public sector, and developers.

Performance Claims and Competitive Positioning

The benchmark numbers have not been independently verified. Reflection says Beam scores on par with GLM-5.2 on advanced reasoning tests and outperforms today's leading Western open models while using "3-4x less inference compute."

The competitive landscape Beam enters is crowded. On the closed side sit Anthropic and OpenAI; on the open side are Chinese developers like DeepSeek, Qwen, and Z.ai, plus Western players including Mistral, Meta, and Cohere. The most direct U.S. comparison may be Inkling, the open model from Mira Murati's Thinking Machines Lab released in July. Reflection's own benchmarks show Beam outscores Inkling on four coding tests where both report results, though Inkling is multimodal and Beam is text-only.

Funding, Compute, and the AI Factory Bet

Reflection has raised roughly $4.7 billion from backers including Nvidia, Sequoia Capital, and Lightspeed Venture Partners, according to PitchBook. Its last round valued the company at a $25 billion pre-money valuation.

The startup has also secured significant compute. This summer, it signed deals collectively worth more than $7 billion with SpaceX and Nebius to access Nvidia's GB300 chips through 2029.

Reflection is targeting enterprises and sovereign nations with a concept it calls "AI factories" — systems that let institutions train Reflection's models on their own proprietary data to build customized, local AI. Nvidia CEO Jensen Huang has long championed this idea, and Nvidia's backing of Reflection aligns with that vision. Axios reported that hedge funds and trading firms are among those interested in such systems. Reflection has already begun testing a sovereign AI factory partnership with South Korea's Shinsegae Group.

What to Watch

Reflection says it will release Beam's weights and full technical details this month, with distribution through hyperscalers and neoclouds and integrations across open source libraries at launch. The company did not respond to TechCrunch's requests for more information in time for publication. The key question is whether independent benchmarks confirm Reflection's efficiency claims — and whether Western enterprises will adopt an open-weight model over established Chinese alternatives.