Google (DeepMind) Gemini 3.8 Flash 89.4% deepmind.google for Gemini 3.8 Flash on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on Terminal-Bench 2.1 30004 Meta AI (originally Facebook AI Research) Muse Spark 1.3 (max)88.8% dev.meta.ai for Muse Spark 1.3 (max) on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported September 2, 2026 Model page Meta Self-reported Record for Muse Spark 1.3 on Terminal-Bench 2.1 80005 Google (DeepMind) Gemini 3.7 Flash 85.8% deepmind.google for Gemini 3.7 Flash on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported August 13, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.7 Flash on Terminal-Bench 2.1 30010 Alibaba (Tongyi) Qwen3.8-Max 86.6% huggingface.co for Qwen3.8-Max on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported August 2, 2026 Model card Qwen Self-reported Qwen Team Record for Qwen3.8-Max on Terminal-Bench 2.1 130008 Google (DeepMind) Gemini 3.6 Flash 78.0% deepmind.google for Gemini 3.6 Flash on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported July 21, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.6 Flash on Terminal-Bench 2.1 30007 Moonshot AI Kimi K3 (max)88.3% huggingface.co for Kimi K3 (max) on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported July 16, 2026 Model card Moonshot AI Self-reported Record for Kimi K3 on Terminal-Bench 2.1 120003 SpaceXAI Grok 4.5 83.3% x.ai for Grok 4.5 on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported July 16, 2026 Launch post SpaceXAI Self-reported Record for Grok 4.5 on Terminal-Bench 2.1 40001 OpenAI GPT-5.6 Luna 84.7% openai.com for GPT-5.6 Luna on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported July 9, 2026 Launch post OpenAI Self-reported Record for GPT-5.6 Luna on Terminal-Bench 2.1 10009 OpenAI GPT-5.6 Sol 88.8% openai.com for GPT-5.6 Sol on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported July 9, 2026 Launch post OpenAI Self-reported Record for GPT-5.6 Sol on Terminal-Bench 2.1 10010 OpenAI GPT-5.6 Terra 87.4% openai.com for GPT-5.6 Terra on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported July 9, 2026 Launch post OpenAI Self-reported Record for GPT-5.6 Terra on Terminal-Bench 2.1 10011 NVIDIA NVIDIA Nemotron 3 Ultra 550B-A55B 56.4% huggingface.co for NVIDIA Nemotron 3 Ultra 550B-A55B on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported June 4, 2026 Model card NVIDIA Self-reported Record for NVIDIA Nemotron 3 Ultra 550B-A55B on Terminal-Bench 2.1 60003 Anthropic Claude Opus 4.8 (adaptive thinking, max)74.6% www-cdn.anthropic.com for Claude Opus 4.8 (adaptive thinking, max) on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported May 28, 2026 System card Anthropic Self-reported Record for Claude Opus 4.8 on Terminal-Bench 2.1 20011 Google (DeepMind) Gemini 3.5 Flash 76.2% deepmind.google for Gemini 3.5 Flash on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported May 19, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.5 Flash on Terminal-Bench 2.1 30006