Terminal-Bench-LILT - Turkish: leaderboard

Metric: Pass rate (%; share of 150 trials, five per task, on the 30 Turkish-language terminal coding tasks written by native-speaker programmers, each verified by its task tests; terminus-2 harness, provider-default reasoning; failed trials caused by infrastructure were re-run). Source: arxiv.org. Saturation forecast: Around December 2026. 3 models tracked.

Top models

#ModelScore
1GPT-5.569.3
2Gemini 3.5 Flash69.3
3Claude Opus 4.866.7

Interactive version: theaggregate.ai/benchmark?slug=terminal-bench-lilt-turkish · How It Works · Data refreshed daily, snapshot 2026-09-29.