Benchmarks › AA-LCR
88.7top score · Kimi K3
273models ranked
173open-weights models
Pass@1higher is better
Oct 20, 2026last updated
How the test works
1
Task
100 tasks
- What the model doesAnswer questions by reasoning across long documents
- Kinds of tasksLong-document reasoning
- DomainsCompany Reports, Industry Reports, Government Consultations, Academia, Legal, Marketing Materials, Survey Reports
2
Run
3 runs per task
- Token budgetApproximately 100k input tokens; minimum 128K context window
- Temperature0 for non-reasoning; 0.6 for reasoning, unless the model lab recommends otherwise
3
Score
Pass@1 (solved on the first try)
- Graded byGPT-5.6 Luna (medium) equality checker
- Better isHigher
Sources: artificialanalysis.ai
AA-LCR 1.1 leaderboardtop 10 of 273
Full leaderboard with filters →| Model | AA-LCR 1.1 | Per task | Price | Speed | Supports | Context |
|---|---|---|---|---|---|---|
88.7max | $2.00 | $3 / $15 | 112 | 1M | ||
88.3thinking | $1.03 | $1 / $2.70 | 93 | 1M | ||
86.3thinking | $0.13 | $0.43 / $0.87 | 54 | 1M | ||
85.3max | $7.63 | $10 / $50 | 69 | 1M | ||
84.7max | $5.98 | $4 / $20 | 96 | 1M | ||
84.3xhigh | $2.63 | $5 / $30 | 113 | 1.05M | ||
84medium | $0.93 | $0.75 / $3.75 | — | 1M | ||
84max | $1.99 | $4 / $20 | 114 | 1.05M | ||
84low | $0.13 | $2 / $10 | 43 | 1.05M | ||
83.7max | $0.18 | $0.20 / $1.20 | 178 | 1.05M |
AA-LCR questions
What is AA-LCR?›
Answer questions by reasoning across long documents. It has 100 tasks. Created by Artificial Analysis.
Which model is best on AA-LCR?›
Kimi K3 leads AA-LCR 1.1 with 88.7, ahead of Step 5 Preview (88.3) and MiMo v2.6 Pro (86.3).
What is the best open-weights model on AA-LCR?›
Kimi K3 with 88.7, then Step 5 Preview (88.3).
How is AA-LCR scored?›
Pass@1 (solved on the first try); gpt-5.6 luna (medium) equality checker; 3 runs per task per task.