Reflection AI unveils Beam, a 501B-parameter (23B active) Apache-2.0 open-weight MoE; weights due later in October
On Oct 5, 2026 Reflection AI announced Beam, its first model: a sparse Mixture-of-Experts with 501B total and 23B active parameters, trained on 23.8T tokens, for coding, reasoning and agentic work. It will be released under Apache 2.0 with weights and a tech report "later in October"; for now it is in early access through Reflection's beta API. Reflection says it matches Z.ai's GLM-5.2 on reasoning with 3-4x less inference compute and beats Thinking Machines' Inkling and Nvidia's Nemotron 3 Ultra, but its own table shows it trailing GLM 5.3, Kimi K3, Qwen 3.8 Max and DeepSeek V4.1 Flash on most tests.
Key facts
- Architecture: sparse MoE, 501B total / 23B active parameters; 256K context in pretraining, extended to 1M in midtraining (API beta lists 256K context, 128K max output)
- Data and compute: 23.8T tokens; pretraining on 6,144 Nvidia GB300 GPUs in under 4 weeks; RL on 10.5K GB300 GPUs for 4 weeks, >100M rollouts, ~1M coding/agentic/STEM environments, ~1.3B sandboxes
- License: Apache 2.0; weights + tech report 'later in October 2026' after final red-teaming; early access via platform.reflection.ai
- API (beta): model id Beam-501B-A23B; endpoints https://api.reflection.ai/v1 and OpenAI-compatible https://api.reflection.ai/openai/v1; knowledge cutoff June 30, 2026; reasoning effort default medium; price not published
- Own benchmarks: SWE-bench Verified 80.9, SWE-Bench Pro v1 65.5, Terminal Bench v2.1 80.1, GPQA Diamond 90.5, AIME 2026 97.8, HLE (no tools) 36.2, BrowseComp (w/ context mgmt) 77.4, MCP Atlas 78.7
- Comparisons in Reflection's table: ahead of Inkling and Nemotron 3 Ultra on most coding tests; behind GLM 5.3, Kimi K3 and DeepSeek V4.1 Flash on Terminal Bench (88.2/88.3/90.6), HLE and SciCode
- Founders: CEO Misha Laskin and CTO Ioannis Antonoglou (ex-DeepMind); Reflection has raised ~$4.6B at a $25B pre-money valuation and holds >$7B in SpaceX and Nebius GB300 compute deals
- Reception: HN thread (78 points within the first hour) criticised the announcement without weights and charts that omit the stronger GLM 5.3 and DeepSeek V4.1 Flash
What happened
A day after Axios reported that a release was imminent, Reflection AI published "Introducing Beam" on Oct 5, 2026. Beam is the lab's first model and is trained from scratch (not distilled, per the founders' interview with Sources). It is a sparse MoE with 501B total and 23B active parameters, aimed at coding, reasoning and agentic workloads. It was trained on 23.8T tokens, with pretraining on 6,144 GB300 GPUs in under four weeks and a four-week RL run on 10.5K GB300 GPUs that Reflection calls "one of the largest scale RL runs conducted by any open lab to date."
The weights are not out yet. Reflection says Beam is in "final red-teaming and evaluations" and that the weights and a technical
report will follow later in October under Apache 2.0. Until then a waitlisted early version runs on Reflection's beta API
(Beam-501B-A23B, OpenAI-compatible endpoint). Distribution through hyperscalers, neoclouds and open-source libraries is planned
(TechCrunch).
Reflection's pitch is efficiency: GLM-5.2-level reasoning at 3-4x less inference compute (about a quarter of GLM-5.3's, per Sources), and "approaching" Qwen 3.8-Max on coding and agentic tasks (Semafor). The blog's own table is mixed. Beam leads the Western open models it lists (Thinking Machines' Inkling, Nvidia's Nemotron 3 Ultra) on most coding benchmarks, but it trails GLM 5.3, Kimi K3, Qwen 3.8 Max and DeepSeek V4.1 Flash on Terminal Bench, HLE, SciCode and AutomationBench. Hacker News commenters said the charts leave out those stronger rivals. CEO Misha Laskin compared closed models to "renting an apartment" (Sources).
Why it matters
This is the first model from the best-funded US lab dedicated to open weights (about $4.6B raised, $25B valuation, more than $7B in GPU deals). An Apache-2.0 500B-class US model gives enterprises and governments that avoid Chinese weights a near-frontier option. The benchmarks still put it behind the best Chinese open models, so it narrows the US–China open-weight gap without closing it. Watch for the actual weight release, the tech report and independent evaluations.
Changelog
- 2026-10-05: created (official blog and developer docs; TechCrunch, Semafor, Sources; Bloomberg paywalled, headline only)
Models
- Beam (Beam-501B-A23B) Reflection AI · preview
People
Ioannis Antonoglou Misha Laskin
Related events
- Axios: Nvidia-backed Reflection AI is about to release its first open-weight model, with other Western open-weight models due in October ★★★
- Thinking Machines Lab releases Inkling, its first open-weights model (975B MoE) ★★★★
- Aleph Alpha releases Kolibri, a 78B-parameter (3.5B active) open-weight English–German 'sovereign' MoE model under Apache 2.0 ★★★
- Open-weight models carry a majority of tokens on Vercel's AI Gateway for the first time (56% in August 2026) ★★★
Sources (8)
- officialReflection AI: Introducing Beam (official blog, Oct 5, 2026)
- docsReflection developer docs: Models (Beam-501B-A23B)
- docsReflection developer docs: API overview (beta)
- pressTechCrunch: Reflection debuts Beam, an open-weight AI model to rival Chinese models at lower compute cost
- pressSemafor: Reflection AI unveils an open-source answer to Chinese labs
- pressBloomberg: Nvidia-backed Reflection unveils open AI model taking on China (paywalled)
- pressSources: Reflection's founders on building a DeepSeek of the West (interview)
- discussionHacker News discussion: Beam, Reflection's 501B open-weight model
id: 2026-10-05-reflection-beam-501b-open-weight · updated 2026-10-05 · open in the interactive timeline