long context benchmark
AA-LCR Leaderboard
AA-LCR is an Agent Arena Long Context Reasoning benchmark in the long_context category, reported as a Score ratio.
Updated Sep 5, 2026
Models19
Model coverage19
MetricScore
EvidenceB
AA-LCR Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelME | Score80.00% | Percentile100.00% | Participants19 | EvidenceC | Evaluated |
| Rank02 | ModelTE | Score73.40% | Percentile94.44% | Participants19 | EvidenceC | Evaluated |
| Rank03 | ModelMA | Score71.20% | Percentile88.89% | Participants19 | EvidenceC | Evaluated |
| Rank04 | ModelUP | Score71.00% | Percentile83.33% | Participants19 | EvidenceC | Evaluated |
| Rank05 | ModelMA | Score70.00% | Percentile77.78% | Participants19 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score68.70% | Percentile72.22% | Participants19 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score68.30% | Percentile66.67% | Participants19 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score66.90% | Percentile61.11% | Participants19 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score66.10% | Percentile55.56% | Participants19 | EvidenceC | Evaluated |
| Rank10 | ModelNV | Score65.40% | Percentile50.00% | Participants19 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score63.00% | Percentile44.44% | Participants19 | EvidenceC | Evaluated |
| Rank12 | ModelMI | Score62.00% | Percentile38.89% | Participants19 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score58.50% | Percentile33.33% | Participants19 | EvidenceC | Evaluated |
| Rank14 | ModelNV | Score58.31% | Percentile27.78% | Participants19 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score57.00% | Percentile22.22% | Participants19 | EvidenceC | Evaluated |
| Rank16 | ModelNV | Score52.00% | Percentile16.67% | Participants19 | EvidenceC | Evaluated |
| Rank17 | ModelLA | Score50.60% | Percentile11.11% | Participants19 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score25.60% | Percentile5.56% | Participants19 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score4.70% | Percentile0.00% | Participants19 | EvidenceC | Evaluated |
AA-LCR Highlights
The leading models and scores on this benchmark.
AA-LCR Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for AA-LCR
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is AA-LCR?
What AA-LCR measures and how its scores work.
AA-LCR is an Agent Arena Long Context Reasoning benchmark.
It reports a Score as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- AA-LCR
- Modality
- text
- Primary category
- long context
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about AA-LCR.
Which model scores highest on AA-LCR?
Muse Glimmer-30B is currently ranked first with 80.00%.
What are the top three models on AA-LCR?
The current leaders are Muse Glimmer-30B (80.00%), Hy3 (73.40%), and Mistral Small 4 (71.20%).
Which AA-LCR model has the lowest official input price?
Mistral Small 4 has the lowest matched official input price at $0.15 input / $0.60 output per 1M tokens.
Which models are fastest among AA-LCR results?
The fastest matched records are Mistral Small 4 (163.71 tok/s via Mistral AI) and MiniMax M2.1 (100.00 tok/s via MiniMax).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.