{"version":1,"type":"story","url":"https://digestai.news/story/reflection-releases-beams-open-weights-under-apache-2-0-license","json":"https://digestai.news/story/reflection-releases-beams-open-weights-under-apache-2-0-license.json","markdown":"https://digestai.news/story/reflection-releases-beams-open-weights-under-apache-2-0-license.md","slug":"reflection-releases-beams-open-weights-under-apache-2-0-license","headline":"Reflection releases open-weight Beam 1.0 with 501B parameters and 1M-token context","summary":"Reflection announced Beam 1.0, an open-weight language model with 501 billion parameters and 23 billion active parameters using a sparse mixture-of-experts architecture. The model supports a 1 million-token context window and is optimized for English, with instruction-tuned versions in BF16, FP8, and NVFP4 formats. Beam claims token-efficient reasoning, outperforming leading open models with three times fewer inference tokens, and excels in software engineering, agentic tool use, and advanced reasoning tasks like math and science problem-solving.\n\nThe model was trained on diverse datasets, including web content, source code, technical documentation, and proprietary vendor data. Reflection emphasizes its safety evaluations, including Open Safety Framework tests and its in-house Iris Safety Evaluations (ISE). Beam is licensed under Apache 2.0 and intended for commercial and research use, with explicit permissions for fine-tuning, distillation, and security research. The model has already been downloaded 440 times on Hugging Face since its release.","keyPoints":["Beam 1.0 has 501B total parameters and 23B active parameters with sparse MoE architecture","Supports 1M-token context window and optimized for English with instruction-tuned versions in BF16/FP8/NVFP4","Achieves 81.6% on SWE-bench Verified, 90.6% on Diamond GPQA, and 80.1% on Terminalbench 2.1"],"whyItMatters":"Beam’s efficiency in token usage and strong performance in coding, reasoning, and agentic tasks could lower barriers for researchers and developers working with large models. Its open-weight release and permissive license may accelerate innovation in AI agents and automation.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Reflection"],"models":["Beam 1.0","Beam-1.0-501B-A23B"],"people":[]},"firstPublishedAt":"2026-10-16T14:22:08Z","updatedAt":"2026-10-16T15:23:10Z","sourceCount":4,"hasPrimarySource":true,"sources":[{"outlet":"huggingface.co","title":"reflection/Beam-1.0-501B-A23B · Hugging Face","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","publishedAt":"2026-10-16T13:23:33Z","type":"primary","primary":true,"lead":true},{"outlet":"AWS Machine Learning Blog","title":"Introducing Reflection Beam: an open-weight model for agentic coding on AWS","url":"https://aws.amazon.com/blogs/machine-learning/introducing-reflection-beam-an-open-weight-model-for-agentic-coding-on-aws","publishedAt":"2026-10-16T15:23:10Z","type":"primary","primary":true,"lead":false},{"outlet":"Unite.AI","title":"Reflection Releases Beam’s Open Weights Under Apache 2.0 License","url":"https://unite.ai/reflection-releases-beams-open-weights-under-apache-2-0-license","publishedAt":"2026-10-16T14:22:08Z","type":"press","primary":false,"lead":false},{"outlet":"huggingface.co","title":"Beam 1.0 - a reflection Collection","url":"https://huggingface.co/collections/reflection/beam-10","publishedAt":"2026-10-16T13:04:58Z","type":"primary","primary":true,"lead":false}],"sourceNotes":null,"discussions":[{"site":"Reddit","url":"https://www.reddit.com/r/LocalLLaMA/comments/1x7g4k3/reflectionbeam10501ba23b_hugging_face/","points":null},{"site":"Reddit","url":"https://www.reddit.com/r/LocalLLaMA/comments/1x7fpaq/beam_10_a_reflection_collection/","points":null}],"thread":{"title":"Reflection AI's Beam Model Saga","url":"https://digestai.news/thread/reflection-unveils-beam-an-open-model-it-says-rivals-glm-5-2-on-reasoning-while","storyCount":2},"cite":{"text":"Digest AI, \"Reflection releases open-weight Beam 1.0 with 501B parameters and 1M-token context\", 16 October 2026, https://digestai.news/story/reflection-releases-beams-open-weights-under-apache-2-0-license","publisher":"Digest AI","title":"Reflection releases open-weight Beam 1.0 with 501B parameters and 1M-token context","datePublished":"2026-10-16T14:22:08Z","url":"https://digestai.news/story/reflection-releases-beams-open-weights-under-apache-2-0-license"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}