---
format: "aidr-story-markdown/v1"
id: "a2c38b693ccc3dfbe904d6afb1797528c8e2c30d51649d397336b8ede5870f06"
canonical_url: "https://aidr.today/a2c38b69?lang=en"
title: "Reflection AI Introduces Beam: A 501B Open-Weight MoE Model With 23B Active Parameters for Coding and Agentic Workloads"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-05T21:04:20.000Z"
category: "Open Source"
topics: ["llm","open-source"]
source_urls: ["https://www.marktechpost.com/2026/10/05/reflection-ai-introduces-beam-a-501b-open-weight-moe-model-with-23b-active-parameters-for-coding-and-agentic-workloads/","https://www.latent.space/p/ainews-reflection-beam-501b-a23b"]
summary: "Reflection AI introduced Beam, a 501B open-weight sparse Mixture-of-Experts model with 23B active parameters for coding and agentic workloads, with Apache 2.0 weights due later in October 2026."
---

# Reflection AI Introduces Beam: A 501B Open\-Weight MoE Model With 23B Active Parameters for Coding and Agentic Workloads

> [Open the canonical story](<https://aidr.today/a2c38b69?lang=en>)

**Published:** 2026-10-05T21:04:20.000Z
**Category:** Open Source
**Topics:** llm, open\-source

## Summary

Reflection AI introduced Beam, a 501B open\-weight sparse Mixture\-of\-Experts model with 23B active parameters for coding and agentic workloads, with Apache 2\.0 weights due later in October 2026\.

## Sources

- [Story source](<https://www.marktechpost.com/2026/10/05/reflection-ai-introduces-beam-a-501b-open-weight-moe-model-with-23b-active-parameters-for-coding-and-agentic-workloads/>)
- [Story source](<https://www.latent.space/p/ainews-reflection-beam-501b-a23b>)