模型档案库
当前快照中出现的全部模型,按最佳榜单成绩展示。不同榜单的分数不可直接比较。
共 123 个模型
Apertus 70B Instruct88.89Artificial Analysis Openness IndexApertus 8B Instruct88.89Artificial Analysis Openness IndexK2 Think V288.89Artificial Analysis Openness IndexK2-V288.89Artificial Analysis Openness IndexMolmo 7B-D88.89Artificial Analysis Openness IndexOlmo 3 7B Instruct88.89Artificial Analysis Openness IndexOlmo 3 7B Think88.89Artificial Analysis Openness IndexOlmo 3.1 32B Instruct88.89Artificial Analysis Openness IndexOlmo 3.1 32B Think88.89Artificial Analysis Openness Indexclaude-fable-5.185.69LiveBench Overallgpt-6-astra85.64LiveBench Overallclaude-opus-5.585.14LiveBench Overallgpt-6.1-sol84.07LiveBench OverallMiniCPM5-1B (Non-reasoning)83.33Artificial Analysis Openness IndexMiniCPM5-1B (Reasoning)83.33Artificial Analysis Openness IndexNemotron 3 Ultra 550B A55B (Reasoning)83.33Artificial Analysis Openness IndexNemotron 3.5 Lightning83.33Artificial Analysis Openness IndexNemotron Cascade 2 30B A3B83.33Artificial Analysis Openness IndexNVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)83.33Artificial Analysis Openness IndexNVIDIA Nemotron 3 Nano 30B A3B (Reasoning)83.33Artificial Analysis Openness IndexNVIDIA Nemotron 3 Nano 4B83.33Artificial Analysis Openness Indexclaude-sonnet-5.580.04LiveBench Overallgemini-4-argon79.30LiveBench Overallgpt-6.1-terra77.79LiveBench Overallmuse-wonder-1.074.15LiveBench Overalldeepseek-v4.1-flash-101772.87LiveBench OverallGranite 4.2 30B72.22Artificial Analysis Openness IndexGranite 4.2 3B72.22Artificial Analysis Openness IndexGranite 4.2 8B72.22Artificial Analysis Openness IndexINTELLECT-3 (Based on GLM-4.5-Air)72.22Artificial Analysis Openness IndexNVIDIA Nemotron Nano 9B V2 (Non-reasoning)72.22Artificial Analysis Openness Indexkimi-k3.1-pro71.29LiveBench Overallgrok-4.770.14LiveBench Overallclaude-haiku-5.569.85LiveBench Overallmimo-v2.6-pro68.48LiveBench Overallgpt-5.6-terra67.79LiveBench Overallgrok-4.667.75LiveBench Overallminimax-m3.1-flash66.89LiveBench Overallgpt-6-luna64.59LiveBench Overallkimi-k364.57LiveBench Overallmistral-large-4-063.49LiveBench Overallmuse-spark-1.363.03LiveBench Overallglm-5.3-flash62.07LiveBench Overallqwen3.8-max61.56LiveBench OverallGemini 4 Carbon60.03Artificial Analysis Intelligence Index v5.0.1gemini-3.8-flash59.21LiveBench Overallstep-5-preview58.61LiveBench Overallqwen3.8-27b58.38LiveBench Overallmuse-spark-open51.63LiveBench Overallsmaug-nemotron-3.5-nvfp450.31LiveBench Overallnemotron-super-vl-3.550.23LiveBench OverallMuse Wonder Full50.08Artificial Analysis Intelligence Index v5.0.1inkling47.05LiveBench Overallnemotron-super-vl-3.5-nvfp446.50LiveBench Overallmistral-large-443.90LiveBench OverallMuse Wonder43.57Artificial Analysis Intelligence Index v5.0.1gemma-4-31b38.72LiveBench OverallTencent Hy434.56Artificial Analysis Intelligence Index v5.0.1Qwen3.8 Max (0902)33.65Artificial Analysis Intelligence Index v5.0.1GLM-5.331.65Artificial Analysis Intelligence Index v5.0.1Inkling 2 Small30.63Artificial Analysis Intelligence Index v5.0.1Ling 3.1 Flash 101929.57Artificial Analysis Intelligence Index v5.0.1K2 Horizon 375B A23B17.82Artificial Analysis Intelligence Index v5.0.1Beam 1.017.73Artificial Analysis Intelligence Index v5.0.1Nemotron 3.5 Super VL13.62Artificial Analysis Intelligence Index v5.0.1Qwen3.8 2.4T A95B93.5%GPQA DiamondMiniMax-M392.9%GPQA DiamondDeepSeek V4 Pro 081392.8%GPQA DiamondClaude Fable 5 (Max, Opus 4.8 Fallback)92.6%GPQA DiamondQwen3.7 Max92.3%GPQA DiamondQwen3.8-Flash-Next92.3%GPQA DiamondGrok 4.383.3%IFBenchClaude Opus 4.783.2%Terminal-Bench v2.1GPT-6 Sol83.0%MMMU-ProGrok 4.20 0309 (Reasoning)82.9%IFBenchGrok 4.20 0309 v2 (Reasoning)81.2%IFBenchMuse Spark80.5%MMMU-ProMiMo-V2.5-Pro (Reasoning)79.9%IFBenchNova 2.0 Pro Preview79.6%IFBenchDeepSeek V4 Flash 042079.2%IFBenchQwen3.5 397B A17B (Reasoning)78.8%IFBenchGemini 3 Flash Preview (Reasoning)78.0%IFBenchQwen3.7 Plus78.0%IFBenchGPT-5.2 Codex77.6%IFBenchGemini 3.1 Flash-Lite77.2%IFBenchQwen3.6 Max Preview76.6%IFBenchDeepSeek V4 Pro 042476.5%IFBenchGLM-5.1 (Reasoning)76.3%IFBenchKimi K2.6 (Reasoning)76.0%IFBenchGPT-5.4 nano75.9%IFBenchMiniMax-M2.775.7%IFBenchQwen3.5 122B A10B (Reasoning)75.7%IFBenchGemma 4 31B (Reasoning)75.6%IFBenchClaude Opus 573.7%DeepSWE v1.1GPT-5.6 Sol72.7%DeepSWE v1.1claude-fable-569.9%DeepSWE v1.1GPT-5.6 Luna67.2%DeepSWE v1.1GPT-5.567.0%DeepSWE v1.1Gemini 3.7 Flash65.5%DeepSWE v1.1deepseek-v4-pro62.8%DeepSWE v1.1Claude Opus 4.859.0%DeepSWE v1.1muse-spark-1-254.9%DeepSWE v1.1Claude Opus 4.7 (Non-reasoning, High)54.5%Terminal-Bench HardClaude Sonnet 553.8%DeepSWE v1.1Grok 4.553.8%DeepSWE v1.1deepseek-v4-flash53.3%DeepSWE v1.1muse-spark-1-153.3%DeepSWE v1.1GPT-5.3 Codex53.0%Terminal-Bench HardGPT-5.4 mini52.3%Terminal-Bench HardGPT-5.451.8%DeepSWE v1.1KAT Coder Pro V249.2%Terminal-Bench HardClaude Opus 4.6 (Non-reasoning, High)48.5%Terminal-Bench HardClaude Opus 4.5 (Reasoning)47.0%Terminal-Bench HardGPT-5.247.0%Terminal-Bench HardGemini 3.6 Flash46.7%DeepSWE v1.1GPT-5.145.5%Terminal-Bench HardQwen3.6 Plus43.9%Terminal-Bench HardGLM-5.243.8%DeepSWE v1.1Gemini 3.5 Flash36.1%DeepSWE v1.1GPT-5.5 Instant (June 2026)31.7%MLCR OverallKimi K2.7 Code30.5%DeepSWE v1.1Claude Sonnet 4.629.9%DeepSWE v1.1Gemini 3.1 Pro Preview11.7%DeepSWE v1.1