MageBench Season 1: leaderboard

Magic: The Gathering matches between LLM agents on the XMage rules engine (28,000+ cards), 1v1 duels and multiplayer Commander, ranked by Elo; Season 1, run by Gregor Stocks, was won by Gemini 3 Pro.

Metric: Rating (imported). Source: github.com. Source: benchmarklist.com. 34 models tracked.

Top models

#ModelScoreOverall rank
1Claude Opus 4.6 (Medium)1747#63 (Claude Opus 4.6)
2GPT-5.2 (Medium)1737#110 (GPT-5.2)
3DeepSeek V3.21682#205
4GPT-5.4 (Medium)1658#79 (GPT-5.4)
5Gemini 3 Flash (Medium)1622#99 (Gemini 3 Flash)
6O3 (Medium)1609#124 (O3)
7Gemini 3.1 Pro (Preview) (Medium)1602#58 (Gemini 3.1 Pro (Preview))
8Qwen 3 Coder 480B A35B1601#279
9Qwen 3 235B A22B1594#313
10Llama 4 Maverick1590#464
11Gemini 3.1 Flash Lite1578#191
12Gemini 2.5 Flash (Medium)1572#245 (Gemini 2.5 Flash)
13Mistral Medium 3.11569#408
14Kimi K2 09051558#288
15Gemini 2.5 Pro (Medium)1540#152 (Gemini 2.5 Pro)

Interactive version: theaggregate.ai/benchmark?slug=magebench-season-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.