BenchGecko Score: leaderboard

Metric: Composite Score. Source: benchgecko.ai. 20 models tracked.

Top models

#ModelScore
1GPT-5.5 Pro99.9
2Claude Mythos Preview97.8
3Qwen 3.5 397B A17B96.3
4DeepSeek V3.2 Speciale95.2
5Claude Fable 593.2
6GLM-5.292.7
7Step 3.5 Flash89.5
8GPT-5 Chat89
9GPT-5.4 Pro (xHigh)87.9
10DeepSeek V4 Pro86.5
11DeepSeek V286.1
12GPT-5.1 Codex Max85.6
13DeepSeek R1 Distill Qwen 14B85.4
14Claude Instant84.6
15Claude Opus 4.882.8

Interactive version: theaggregate.ai/benchmark?slug=benchgecko-score · How It Works · Data refreshed daily, snapshot 2026-09-05.