AIEvals

Benchmarks

LongBench v2

Answering questions that require understanding and reasoning over long input contexts.

Long-context reasoningLongBench, reported by Reflection AIVendor-reported onlyReported by Reflection AI

How it is run

Reflection AI reports four model scores on LongBench v2, including Beam at 65.5%.

What it cannot tell you

The checked source does not specify reasoning effort, task selection, context cap or attempt count. These results remain a publisher table.

Why there is no independent table

No independent suite in this dataset runs LongBench v2. The numbers below are the ones model publishers report about their own models, which is a different kind of evidence: each publisher chooses its own harness and effort level, so a row is readable on its own and not against the other rows.

Read this with

Benchmarks in the same category, which usually disagree with this one on purpose.

  • AA-LCR (version not specified)Same category. Finding relevant information and reasoning about it across long input contexts.
  • AA-LCR v1.1Same category. Extracting, reasoning about and synthesizing information in long documents.

As reported by each publisher

Kept separate per publisher, because the settings are not shared.

Reflection AI

Introducing Beam: model capability comparisons · 5 October 2026

RankModelScore
2Beam65.5%
4Nemotron 3 Ultra61.9%
3GLM 5.264.0%
1Qwen3.8 Max66.3%

Ranks compare reported scores in the same published row; ties share a rank.

Reflection AI’s four capability tables, read 6 October 2026. All 21 rows and eight model columns are transcribed; NR means not reported and remains blank. The source’s Qwen 3.8 Max column uses this catalogue’s Qwen3.8 Max spelling. Beam results are publisher reports; Reflection AI cites Artificial Analysis and Datacurve for comparison-model evaluations. The announcement does not specify a common harness, effort or trial count for every row. Ranks are derived within each published row from reported values only, with ties sharing a rank; they do not establish an independent or overall model ranking. AutomationBench uses the public split; AA-Omniscience is an index on the public split only, calculated by Reflection AI. BrowseComp and DeepSearchQA use context management. AA-LCR has no stated version. These settings remain separate from other source fields.

reflection.ai/blog/introducing-beam

Sources

Every page used for this benchmark, read on 17 October 2026.