AI Tools Review

Beam 1.0

By Reflection

Released: 2026-10-16

LLM
Reflection
Open Weights
Free
New

Beam 1.0 is Reflection's open-weight AI model, released 16/10/2026. Reflection AI's launch post presents Beam as a text-only sparse mixture-of-experts model with 501B total and 23B active parameters, 23.8T pretraining tokens and an Apache 2.0 weight release promised for later in October 2026. The post says it matches GLM-5.2 on advanced reasoning with 3 to 4 times less inference compute (FLOPs approximated as 2 x active parameters x mean generated tokens). Tables reflect an update dated 27 October 2026 and the page gives no evaluation settings, so rows carry no setting. Entries the page marks NR are omitted from the comparison columns.

Visit Beam 1.0

AI-Powered

Leverages advanced AI technology to deliver cutting-edge capabilities and results.

Fast & Efficient

Optimized performance ensures quick results without compromising on quality.

Purpose-Built

Specifically designed for llms tasks and workflows.

How does Beam 1.0 (Beam 501B) score on independent benchmarks?

Beam 1.0 (Beam 501B) ranks #27 of 28 models on the Artificial Analysis Intelligence Index (v5.0.1) with a score of 17.7, as measured on 22 November 2026.

Intelligence Index
17.7
AA v5.0.1
Coding Agent Index
15.5
in Mirror
Cost per task
$0.28
AA Intelligence Index task
Price per 1M tokens
$0.4 / $1.2
input / output

Where does Beam 1.0 (Beam 501B) rank against other models?

How good is Beam 1.0 (Beam 501B) at agentic coding?

How does Beam 1.0 (Beam 501B) do on individual tests?

Evaluation (run by Artificial Analysis)Beam 1.0 (Beam 501B)Best of any model
Terminal-Bench 4.03%67.2%
FrontierCode 2.05.5%48.8%
Humanity's Last Exam34.7%74.3%
GDPval-AA (Elo)11771866

How has Beam 1.0 (Beam 501B)'s score changed over time?

We have tracked Beam 1.0 (Beam 501B) on Intelligence Index v5.0.1 since 22/11/2026. A point is added each day the score, coding index or price changes, so this chart fills in over time.

What does the Beam 1.0 (Beam 501B) system card report?

Reflection AI's launch post presents Beam as a text-only sparse mixture-of-experts model with 501B total and 23B active parameters, 23.8T pretraining tokens and an Apache 2.0 weight release promised for later in October 2026. The post says it matches GLM-5.2 on advanced reasoning with 3 to 4 times less inference compute (FLOPs approximated as 2 x active parameters x mean generated tokens). Tables reflect an update dated 27 October 2026 and the page gives no evaluation settings, so rows carry no setting. Entries the page marks NR are omitted from the comparison columns.

These are Reflection AI's own published results (self-reported, with the vendor's harness and settings), so they are not directly comparable with other labs' numbers. Source: Introducing Beam (tables updated 27 October 2026) (05/10/2026). Last checked 22/11/2026.

Coding

BenchmarkBeam 1.0 (Beam 501B)GLM 5.2GLM 5.3Kimi K3
DeepSWE v1.144.4%44%61%68%
SWE Bench Pro v2-Hard77.2%n/a84.3%88.2%
SWE Bench Pro v165.5%62.1%n/an/a
Terminal Bench v2.180.1%81%88.2%88.3%
SWE Atlas Codebase QnA34.6%n/a61%68%
SWEBench Multilingual78%n/an/an/a
SWEBench Verified80.9%n/an/an/a
SciCode49.7%n/a59%58.7%

Reasoning & maths

BenchmarkBeam 1.0 (Beam 501B)GLM 5.2GLM 5.3Kimi K3
AIME 202697.8%99.2%n/an/a
CritPt AA16.3%20.9%19.1%23.4%

Knowledge & science

BenchmarkBeam 1.0 (Beam 501B)GLM 5.2GLM 5.3Kimi K3
HLE (no tools)36.2%40.5%42.3%46.9%
HLE (with tools)51.2%54.7%62.5%n/a
GPQA Diamond90.5%91.2%91.7%93.5%
AA Omniscience (index), public split onlyReflection's own results on the public split13 pointsn/a22.6n/a

Agents & tool use

BenchmarkBeam 1.0 (Beam 501B)GLM 5.2GLM 5.3Kimi K3
AutomationBench public37%26.2%48.2%46.7%
MCP Atlas78.7%77.8%84.2%82.3%
tau3 banking38%37.1%n/a37.1%
BrowseComp with context management77.4%n/an/a91.2%
DeepSearchQA with context management80.1%n/an/a95%

Long context

BenchmarkBeam 1.0 (Beam 501B)GLM 5.2GLM 5.3Kimi K3
AA-LCR79.3%78.3%79.7%88.7%
LongBench v266.9%64%n/an/a

Other

BenchmarkBeam 1.0 (Beam 501B)GLM 5.2GLM 5.3Kimi K3
IFBench79.7%73.3%n/an/a

What did safety testing find for Beam 1.0 (Beam 501B)?

Safety level: Not yet published; the post says Beam is undergoing final red-teaming and evaluations

  • Alignment uses deliberative alignment, with safety policy incorporated into reasoning and an adversarially built dataset, plus safety RL across single-turn, multi-turn, jailbreak and agentic scenarios.
  • A second model was trained with separate SFT and RL pipelines, and two teachers were merged using multi-teacher on-policy distillation.
  • Safety evaluation results are to appear in the technical report, and internal safety evaluations are to be open-sourced; no safety scores are on the page.

Independent figures: Artificial Analysis, refreshed daily on our AI model benchmarks page. Full analysis: our Beam 1.0 (Beam 501B) deep dive.

Reflection Model Timeline

Beam 1.0Current

RL rollouts to 256K tokens; mid-training extends effective context to 1M tokens context

Specifications

pricing$0.4/$1.2 per 1M tokens (input/output)
context WindowRL rollouts to 256K tokens; mid-training extends effective context to 1M tokens