GM League Arena - NBA Elo: leaderboard
GM League Arena (arXiv 2611.05147v1; called GM Bench in the paper) makes each LLM the general manager of one team in a shared simulated NBA league for five seasons (2022–2026), "managing players and assets through selection, trades, and contract renewals". Table 1 Elo is "the arithmetic mean of season-end ratings over the five seasons, with ratings initialized to 1500 at the start of each season", so it is relative to the other 29 teams in this league; a rule-based Native GM team takes part but is not a model row. "Configurable reasoning effort is set to high for every model that supports this setting"; each model has "one complete five-season run" and no confidence intervals are reported. Metric: Mean season-end Elo.
Metric: Mean season-end Elo over five simulated NBA seasons (2022-2026; ratings reset to 1500 each season, K=20, 400-point logistic scale; one shared league of 29 LLM general managers and a rule-based Native GM team; official APIs, high reasoning effort where supported; one full-horizon run per model). Source: arxiv.org. Saturation forecast: Around November 2027. 29 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 | 1977.3 | #3 |
| 2 | GPT-6 Astra | 1907.7 | #5 |
| 3 | GPT-6.1 Sol | 1838.7 | #6 |
| 4 | Claude Sonnet 5.5 | 1813.3 | #14 |
| 5 | Claude Fable 5.1 | 1794.4 | #10 |
| 6 | Grok 4.7 | 1784.3 | #26 |
| 7 | Gemini 3.1 Pro (Preview) | 1753.1 | #65 |
| 8 | Grok 4.6 | 1713.6 | #30 |
| 9 | Kimi K3 | 1698.8 | #35 |
| 10 | DeepSeek V4 Pro | 1666.3 | #112 |
| 11 | GLM-5.3 | 1652.6 | #43 |
| 12 | Qwen 3.8 Max | 1602.5 | #34 |
| 13 | GPT-6 Luna | 1582.6 | #62 |
| 14 | Gemini 3.8 Flash | 1525.7 | #31 |
| 15 | GPT-5.6 Terra | 1499 | #41 |
No result here: #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=gm-league-arena-nba-elo · How It Works · Data refreshed daily, snapshot 2026-11-06.