Reflection AI's Beam Weights Released Under Apache 2.0 With 40+ Partners
A 501B-parameter open-weight model with an Apache 2.0 license is available for enterprises and governments to self-host and fine-tune. The release includes the training harness, enabling others to inspect and replicate aspects of the training process.
Reflection AI released Beam's open weights on October 16, 2026, under the Apache 2.0 license, along with the tokenizer, reference implementation in BF16, FP8, and NVFP4 formats, a DFlash2 drafter, and Mirror CLI, the training harness used to build Beam. The release was done in collaboration with over 40 partners. Beam is a sparse mixture-of-experts model with 501 billion total parameters and 23 billion active parameters per token. Pretraining covered 23.8 trillion curated tokens on 6,144 GB300 NVL72 GPUs in under four weeks. The reinforcement learning phase used roughly 10,500 of the same GPUs for four additional weeks, generating over 100 million rollouts across nearly one million environments and about 1.3 billion sandboxes, with no sign of a plateau. Reflection positions Beam for enterprises and governments seeking to self-host, fine-tune with proprietary data, and avoid dependence on closed APIs. Reflection AI is backed by Nvidia.
Reflection AI released Beam's weights on October 16 under the Apache 2.0 license, making the 501-billion-parameter sparse mixture-of-experts model available for self-hosting and fine-tuning.
The release includes BF16, FP8, and NVFP4 formats, a DFlash2 drafter, and Mirror CLI, the training harness used to build Beam, in collaboration with over 40 partners.
October 12 reporting added a benchmark comparison not included in Reflection AI's headline chart: Beam scores 44.4 on DeepSWE 1.1 against the newer GLM-5.3's 61.0, extending the gap beyond earlier reported measures.
The reporting also clarified that Beam's weights and technical report, from the October 5 launch, remain planned for later in October and have not yet been published.
The Financial Times reported October 10 that Nvidia is in early discussions to acquire Reflection AI or expand its existing $800M stake, with an acqui-hire structure, in which Nvidia would hire Reflection's staff and license its technology without triggering a full antitrust review, as one option under consideration.
The FT said a deal could materialize within weeks and described the Trump administration as supportive of Reflection AI as a competitor to Chinese AI efforts.
Subsequent coverage on October 9 confirmed that Reflection AI is backed by Nvidia, a detail absent from the initial release reports.
Beam, released October 5, uses a sparse mixture-of-experts architecture with 501 billion total parameters and 23 billion active per token, trained on 23.8 trillion tokens and released under Apache 2.0. Reflection claims a three-to-four-times inference compute efficiency advantage over GLM-5.2, though one source notes the comparison uses estimated FLOPs rather than measured serving cost.
Coverage published October 7 framed Beam's market case around compliance and control rather than benchmark performance, citing Reflection CEO Misha Laskin's stated focus on regulated industries and governments that cannot use Chinese AI models.
A source also reported that Beam trails top Chinese open-source models and leading closed systems from OpenAI and Anthropic on benchmarks, placing the model's competitive argument around affordability and control rather than capability.
Reflection AI released Beam on October 5, a 501-billion-parameter sparse mixture-of-experts model with 23 billion active per token under Apache 2.0, pretrained on 23.8 trillion tokens using 6,144 Nvidia GB300 GPUs.
A reinforcement learning phase on 10,500 GPUs generated over 100 million rollouts with no sign of a plateau; Reflection claims three-to-four times inference compute efficiency over GLM-5.2, though one source notes the figure rests on estimated forward-pass FLOPs rather than measured serving cost. On Terminal Bench v2.1, Beam scores 80.1 against GLM-5.2's 81.0, but DeepSeek V4.1 Flash at 90.6 and Kimi K3 at 88.3 do not appear in Reflection's headline benchmark chart.
Analysis published showing GLM-5.2 and Kimi K3 achieve near-frontier performance at 17-35% of leading proprietary model costs
I send one email each morning with the stories that moved. If you would rather just read here, that works too.
Subscribe free