SimAgentsAcross the notebook
Meet the models.
Results are grouped by the names used in the research. A model may appear under different names or settings in different tests; each published name has its own page.
Arcee Trinity Large Thinking1 benchmark ↗Baidu Ernie 5.02 benchmarks ↗Baidu Ernie 5.13 benchmarks ↗ByteDance Seed2.0 Pro3 benchmarks ↗Claude 4.5 Haiku1 benchmark ↗Claude Fable 5 (high reasoning)1 benchmark ↗Claude Fable 5 (high)2 benchmarks ↗Claude Fable 5.1 (high reasoning)1 benchmark ↗Claude Fable 5.1 (high)2 benchmarks ↗Claude Haiku 5.5 (high reasoning)1 benchmark ↗Claude Opus 4.5 (no reasoning)1 benchmark ↗Claude Opus 4.5 Thinking 16K1 benchmark ↗Claude Opus 4.6 (high reasoning)1 benchmark ↗Claude Opus 4.6 (no reasoning)1 benchmark ↗Claude Opus 4.6 Thinking 16K2 benchmarks ↗Claude Opus 4.7 (high reasoning)1 benchmark ↗Claude Opus 4.7 (high)2 benchmarks ↗Claude Opus 4.7 (no reasoning)1 benchmark ↗Claude Opus 4.8 (high reasoning)1 benchmark ↗Claude Opus 4.8 (high)2 benchmarks ↗Claude Opus 4.8 (xhigh reasoning)1 benchmark ↗Claude Opus 4.8 (xhigh)1 benchmark ↗Claude Opus 5 (high reasoning)1 benchmark ↗Claude Opus 5 (high)2 benchmarks ↗Claude Opus 5 (xhigh reasoning)1 benchmark ↗Claude Opus 5 (xhigh)1 benchmark ↗Claude Opus 5.5 (high reasoning)1 benchmark ↗Claude Opus 5.5 (high)2 benchmarks ↗Claude Opus 5.5 (medium reasoning)1 benchmark ↗Claude Sonnet 4.5 (no reasoning)1 benchmark ↗Claude Sonnet 4.5 Thinking 16K1 benchmark ↗Claude Sonnet 4.6 (high reasoning)1 benchmark ↗Claude Sonnet 4.6 (high)1 benchmark ↗Claude Sonnet 4.6 (no reasoning)2 benchmarks ↗Claude Sonnet 4.6 Thinking 16K2 benchmarks ↗Claude Sonnet 4.6 Thinking 32K1 benchmark ↗Claude Sonnet 5 (high reasoning)1 benchmark ↗Claude Sonnet 5 (high)1 benchmark ↗Claude Sonnet 5.5 (high reasoning)1 benchmark ↗DeepSeek V3.23 benchmarks ↗DeepSeek V3.2 (no reasoning)1 benchmark ↗DeepSeek V4 Flash1 benchmark ↗DeepSeek V4 Pro (high reasoning)1 benchmark ↗DeepSeek V4 Pro (high)1 benchmark ↗DeepSeek V4 Pro 0813 (high)1 benchmark ↗DeepSeek V4 Pro Preview3 benchmarks ↗DeepSeek V4.1 Flash (high reasoning)1 benchmark ↗DeepSeek V4.1 Flash (high)2 benchmarks ↗GLM-51 benchmark ↗GLM-5.13 benchmarks ↗GLM-5.2 (high reasoning)1 benchmark ↗GLM-5.2 (max)2 benchmarks ↗GLM-5.3 (high reasoning)1 benchmark ↗GLM-5.3 (high)1 benchmark ↗GLM-5.3 (max)1 benchmark ↗GPT-5.2 (high reasoning)1 benchmark ↗GPT-5.2 (low reasoning)1 benchmark ↗GPT-5.2 (medium reasoning)1 benchmark ↗GPT-5.2 (medium)1 benchmark ↗GPT-5.2 (no reasoning)1 benchmark ↗GPT-5.2 (xhigh reasoning)1 benchmark ↗GPT-5.2 Pro1 benchmark ↗GPT-5.4 (high reasoning)1 benchmark ↗GPT-5.4 (high)1 benchmark ↗GPT-5.4 (medium reasoning)1 benchmark ↗GPT-5.4 (medium)1 benchmark ↗GPT-5.4 (no reasoning)2 benchmarks ↗GPT-5.4 (xhigh reasoning)1 benchmark ↗GPT-5.4 (xhigh)1 benchmark ↗GPT-5.4 Mini (xhigh reasoning)1 benchmark ↗GPT-5.5 (high reasoning)1 benchmark ↗GPT-5.5 (high)1 benchmark ↗GPT-5.5 (medium reasoning)1 benchmark ↗GPT-5.5 (no reasoning)1 benchmark ↗GPT-5.5 (xhigh reasoning)1 benchmark ↗GPT-5.5 (xhigh)1 benchmark ↗GPT-5.6 Luna (high reasoning)1 benchmark ↗GPT-5.6 Sol (high reasoning)1 benchmark ↗GPT-5.6 Sol (high)2 benchmarks ↗GPT-5.6 Sol (xhigh reasoning)1 benchmark ↗GPT-5.6 Sol (xhigh)1 benchmark ↗GPT-5.6 Terra (high reasoning)1 benchmark ↗GPT-6 Astra (high reasoning)1 benchmark ↗GPT-6 Astra (high)2 benchmarks ↗GPT-6 Astra (xhigh reasoning)1 benchmark ↗GPT-6 Luna (high reasoning)1 benchmark ↗GPT-6 Sol (high reasoning)1 benchmark ↗GPT-6.1 Sol (high reasoning)1 benchmark ↗GPT-OSS-120B2 benchmarks ↗Gemini 3 Flash Preview1 benchmark ↗Gemini 3 Pro Preview1 benchmark ↗Gemini 3.1 Flash-Lite Preview2 benchmarks ↗Gemini 3.1 Pro Preview3 benchmarks ↗Gemini 3.5 Flash3 benchmarks ↗Gemini 3.5 Flash-Lite (high reasoning)1 benchmark ↗Gemini 3.6 Flash1 benchmark ↗Gemini 3.7 Flash1 benchmark ↗Gemini 3.7 Flash (high)2 benchmarks ↗Gemini 3.8 Flash (high reasoning)1 benchmark ↗Gemini 3.8 Flash (high)2 benchmarks ↗Gemma 4 31B IT1 benchmark ↗Gemma 4 31B Reasoning2 benchmarks ↗Grok 4.1 Fast Non-Reasoning1 benchmark ↗Grok 4.1 Fast Reasoning1 benchmark ↗Grok 4.20 0309 (Non-Reasoning)2 benchmarks ↗Grok 4.20 0309 (Reasoning)2 benchmarks ↗Grok 4.20 Multi-Agent Exp Beta 03041 benchmark ↗Grok 4.20 Non-Reasoning Exp Beta 03041 benchmark ↗Grok 4.20 Reasoning Exp Beta 03041 benchmark ↗Grok 4.33 benchmarks ↗Grok 4.5 (high reasoning)1 benchmark ↗Grok 4.5 (high)2 benchmarks ↗Grok 4.6 (high reasoning)1 benchmark ↗Grok 4.6 (high)2 benchmarks ↗Grok 4.6 (xhigh reasoning)1 benchmark ↗Grok 4.7 (high reasoning)1 benchmark ↗Grok 4.7 (high)2 benchmarks ↗Kimi K2.5 Thinking3 benchmarks ↗Kimi K2.63 benchmarks ↗Kimi K33 benchmarks ↗Ling 2.6 1T1 benchmark ↗Llama 4 Maverick2 benchmarks ↗LongCat Flash Thinking1 benchmark ↗MiniMax-M21 benchmark ↗MiniMax-M2.11 benchmark ↗MiniMax-M2.51 benchmark ↗MiniMax-M2.73 benchmarks ↗MiniMax-M33 benchmarks ↗Mistral Large 33 benchmarks ↗Mistral Large 4 (high)2 benchmarks ↗Mistral Medium 3.12 benchmarks ↗Mistral Medium 3.51 benchmark ↗Mistral Medium 3.5 (high)2 benchmarks ↗Muse Spark 1.1 (high reasoning)1 benchmark ↗Muse Spark 1.1 (high)2 benchmarks ↗Muse Spark 1.2 (high reasoning)1 benchmark ↗Muse Spark 1.2 (high)1 benchmark ↗Muse Spark 1.3 (high reasoning)1 benchmark ↗Muse Spark 1.3 (high)2 benchmarks ↗Nemotron 3 Super1 benchmark ↗Qwen 3 Max Preview1 benchmark ↗Qwen 3 Max Thinking1 benchmark ↗Qwen 3.6 Max Preview3 benchmarks ↗Qwen 3.6 Plus2 benchmarks ↗Qwen 3.7 Flash1 benchmark ↗Qwen 3.7 Max3 benchmarks ↗Qwen 3.7 Plus1 benchmark ↗Qwen 3.8 Max3 benchmarks ↗Qwen3 Max (2026-01-23)1 benchmark ↗Qwen3.5-122B-A10B1 benchmark ↗Qwen3.5-27B1 benchmark ↗Qwen3.5-397B-A17B2 benchmarks ↗Qwen3.6-35B-A3B1 benchmark ↗Qwen3.8-27B2 benchmarks ↗Step 3.5 Flash1 benchmark ↗Step 3.7 Flash (high reasoning)1 benchmark ↗Step 3.7 Flash (high)1 benchmark ↗Tencent Hy3 (high)1 benchmark ↗Tencent Hy3 (non-thinking)1 benchmark ↗Tencent Hy3 Preview1 benchmark ↗Tencent Hy3 Preview (high)1 benchmark ↗Tencent Hy4 Preview1 benchmark ↗Tencent Hy4 Preview (high)1 benchmark ↗Xiaomi MiMo V2 Pro3 benchmarks ↗Xiaomi MiMo V2.5 Pro3 benchmarks ↗Xiaomi MiMo V2.6 Pro (thinking)2 benchmarks ↗