[whatmodel.sh](../../README.md) / Beam 1.0 501B A23B # Beam 1.0 501B A23B Reflection · released 2026-10 · 501.3B params (23.4B active) · apache-2.0 **score 70.3** ±1.3 · #78 of 732 · from 3 benchmark results across 1 source Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at _{● marks Beam 1.0 501B A23B among every model on that task or test, worst to best.}

task            vs the others            rank  counts  tests
agent tasks     ▁▁▃▆█▄▅▅●▄▂▁      #55 of 313*     48%     10
coding          ▁▁▁▃▅▇█▇▅●▃▁      #79 of 469*     33%      7
chat & writing  ▁▁▁▃▆▄▆█▆▇●▃      #75 of 402*     19%      1
math                           not ranked yet       —      1
web search                     not ranked yet       —      2

agent tasks · #55 of 313 · 48% of its score

Beats [Gemini 3.1 Pro (Preview)](gemini-3-1-pro.md), [Muse Spark 1.1](muse-spark-1-1.md), [Muse Spark 1.2](muse-spark-1-2.md) here, though they rank higher overall.

test                                     result  vs the others in its run    difficulty            counts  why
Arena Agent (October 2026 aggrega…  -0.0493 IPS  ▁▁▁▃●▂█▁▂▂▁▁ beats 14/59    avg -0.379 IPS           48%  sharp
SWE-bench Verified (maker)                81.6%  ▁▁▁▁▁▁▂▄█●▁▁ beats 132/138  hard · avg 19%             —  not counted yet: no other model in the …
SWE-bench Multilingual (maker)            78.3%  ▁··▁▃▁▅█●▂▂▁ beats 52/65    hard · avg 18%             —  not counted yet: no other model in the …
Terminal-Bench 2.1 (maker)                80.1%  ▁·▁·▁▁▄▃▆█●█ beats 43/63    very hard · avg 7.2%       —  not counted yet: no other model in the …
Humanity's Last Exam with tools (…        51.1%  ▁·▁▂▂▁▄▄▇●▄▅ beats 37/62    hard · avg 15%             —  not counted yet: no other model in the …
MCP-Atlas (public set) (maker)            80.8%  ▂·▁▂▂·▁▃█▆▆● beats 38/39    hard · avg 30%             —  not counted yet: no other model in the …
DeepSWE v1.1 (resolved) (maker)           43.3%  ▂····●▁▁▆███ beats 3/31                                —  not counted yet: no other model in the …
τ³-bench Banking (maker)                  36.1%  ▂▂·▅█▂▃▃▃▆▂● beats 23/24    very hard · avg 5.3%       —  not counted yet: no other model in the …
SWE-bench Pro (maker)                     66.1%  ▄▂▂▂▄▂▆█·▂·● beats 16/16                               —  not counted yet: no other model in the …
BrowseComp with context managemen…        76.8%  ▄·▄▂·●▂·▆▄·▂ beats 6/15                                —  not counted yet: no other model in the …

coding · #79 of 469 · 33% of its score

Beats [GPT-5.2](gpt-5-2.md), [GPT-5](gpt-5.md), [Grok 4.20 (0309 reasoning)](grok-4-20.md) here, though they rank higher overall. Trails [Llama 3.1 Nemotron Ultra 253B v1](llama-3-1-nemotron-ultra-253b.md), [MiMo-V2.5 (310B-A15B)](mimo-v2-5.md), [A.X K2 (688B-A33B)](ax-k2.md) here, though they rank lower overall.

test                                result  vs the others in its run    difficulty           counts  why
Arena WebDev (Code Arena)           1436.0  ▁▁▃▂▅●▃▆▅▅▁▂ beats 57/129   avg 981                 33%  sharp
SWE-bench Verified (maker)           81.6%  ▁▁▁▁▁▁▂▄█●▁▁ beats 132/138  hard · avg 19%            —  not counted yet: no other model in the …
SWE-bench Multilingual (maker)       78.3%  ▁··▁▃▁▅█●▂▂▁ beats 52/65    hard · avg 18%            —  not counted yet: no other model in the …
SciCode (level unspecified) (make…   49.2%  ▁····▃▃▄▅█▆● beats 39/40    very hard · avg 13%       —  not counted yet: no other model in the …
LiveCodeBench (version unspecifie…   88.3%  ▁···▁▁▁█▇▃▄● beats 38/38    hard · avg 21%            —  not counted yet: no other model in the …
DeepSWE v1.1 (resolved) (maker)      43.3%  ▂····●▁▁▆███ beats 3/31                               —  not counted yet: no other model in the …
SWE-bench Pro (maker)                66.1%  ▄▂▂▂▄▂▆█·▂·● beats 16/16                              —  not counted yet: no other model in the …

chat & writing · #75 of 402 · 19% of its score

Beats [Claude Haiku 5.5](claude-haiku-5-5.md), [LongCat 2.0 (1.6T-A48B)](longcat-2-0.md), [GPT-5](gpt-5.md) here, though they rank higher overall. Trails [Gemma 4 26B-A4B](gemma-4-26b-a4b.md), [Nemotron Cascade 2 30B-A3B](nemotron-cascade-2-30b-a3b.md), [Gemma 4 31B](gemma-4-31b.md) here, though they rank lower overall.

test            result  vs the others in its run    difficulty  counts  why
LMArena (text)  1435.7  ▁▁▁▂▃▃▃▆▆▇●▃ beats 296/385  avg 1260       19%  blunt

math · 0.0% of its score

test               result  vs the others in its run  difficulty        counts  why
AIME 2026 (maker)   97.7%  ▁▁▁▂▂▂▆█▆●·▁ beats 51/52  medium · avg 48%       —  not counted yet: no other model in the …

web search · 0.0% of its score

test                                result  vs the others in its run  difficulty      counts  why
Humanity's Last Exam with tools (…   51.1%  ▁·▁▂▂▁▄▄▇●▄▅ beats 37/62  hard · avg 15%       —  not counted yet: no other model in the …
BrowseComp with context managemen…   76.8%  ▄·▄▂·●▂·▆▄·▂ beats 6/15                        —  not counted yet: no other model in the …

other tests (5)

test                                result  vs the others in its run    difficulty            counts  why
GPQA Diamond (maker)                 90.6%  ▁▁▁▁▂▄▅▅██●▄ beats 281/322  hard · avg 39%             —  not counted yet: no other model in the …
Humanity's Last Exam (maker)         36.0%  ▁▁▁▄▆▆▇█●▆▁▃ beats 89/111   very hard · avg 3.8%       —  not counted yet: no other model in the …
IFBench (aggregation unspecified)…   79.6%  ▂▆▂▄▄▄▂▄▄▆█● beats 79/86    medium · avg 44%           —  not counted yet: no other model in the …
LongBench v2 (maker)                 67.5%  ▂·▁·▂▃▁▄▄▃█● beats 32/33    very hard · avg 32%        —  not counted yet: no other model in the …
AutomationBench (Public) (maker)     39.3%  █····▄····●▄ beats 3/4                                 —  not counted yet: no other model in the …

## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md)

source                         results  share of the score
LMArena leaderboard dataset          3   99% ▇▇▇▇▇▇▇▇▇▇▇▇
prior: pull toward the middle             1% ▇
weighted average                        = 70.3

_{Licences: LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/))}

Not counted (17)

source                  test                            score  why
Reflection Beam 1.0 m…  DeepSWE v1.1 (resolved)         43.3%  too few models have it yet
Reflection Beam 1.0 m…  SWE-bench Pro                   66.1%  too few models have it yet
Reflection Beam 1.0 m…  Terminal-Bench 2.1              80.1%  too few models have it yet
Reflection Beam 1.0 m…  SWE-bench Multilingual          78.3%  too few models have it yet
Reflection Beam 1.0 m…  SWE-bench Verified              81.6%  too few models have it yet
Reflection Beam 1.0 m…  AIME 2026                       97.7%  too few models have it yet
Reflection Beam 1.0 m…  Humanity's Last Exam            36.0%  too few models have it yet
Reflection Beam 1.0 m…  Humanity's Last Exam with too…  51.1%  too few models have it yet
Reflection Beam 1.0 m…  SciCode (level unspecified)     49.2%  too few models have it yet
Reflection Beam 1.0 m…  GPQA Diamond                    90.6%  too few models have it yet
Reflection Beam 1.0 m…  LiveCodeBench (version unspec…  88.3%  too few models have it yet
Reflection Beam 1.0 m…  AutomationBench (Public)        39.3%  too few models have it yet
Reflection Beam 1.0 m…  MCP-Atlas (public set)          80.8%  too few models have it yet
Reflection Beam 1.0 m…  τ³-bench Banking                36.1%  too few models have it yet
Reflection Beam 1.0 m…  BrowseComp with context manag…  76.8%  too few models have it yet
Reflection Beam 1.0 m…  LongBench v2                    67.5%  too few models have it yet
Reflection Beam 1.0 m…  IFBench (aggregation unspecif…  79.6%  too few models have it yet

## How to run it ### Through an API api (Reflection) · price unknown · [developers.reflection.ai](https://developers.reflection.ai/models) _{[context source](https://huggingface.co/reflection/Beam-1.0-501B-A23B/blob/78f8f654d40a40919b8e9954b40dd879ecb9a692/README.md): Official card: 1M context; exact 1,048,576 limit corroborated by Reflection developer model documentation, observed 2026-10-21.} Model repository and documentation: [huggingface.co/reflection](https://huggingface.co/reflection/Beam-1.0-501B-A23B) [interactive version](https://whatmodel.sh/model/beam-1-0-501b-a23b) · [JSON](https://whatmodel.sh/api?format=json&all=1)