Live benchmark results, sourced from published third-party evaluations. Real pricing and practical context will appear here as they are licensed — see the empty cards below rather than a guess.
“Epoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks' [online resource].” · epoch.ai/benchmarks · Licensed CC BY 4.0 · data last verified 2026-07-26 · Methodology · Data Sources
| Rank | Model | Provider | Overall (ECI) | Coding | Reasoning | Context | Speed | Reference Price |
|---|---|---|---|---|---|---|---|---|
| 1 | GPT-5.6 Sol | OpenAI | 161.7 | — | 0.935 (max) | — | — | — |
| 2 | GPT-5.5 Pro | OpenAI | 160.9 | — | 0.939 (xhigh-pre-release) | — | — | — |
| 3 | Claude Fable 5 | Anthropic | 160.7 | — | — | — | — | — |
| 4 | Claude Opus 5 | Anthropic | 159.4 | — | 0.939 (max) | — | — | — |
| 5 | GPT-5.5 | OpenAI | 158.5 | 0.806 (xhigh-pre-release) | 0.940 (xhigh-pre-release) | — | — | — |
| 6 | GPT-5.6 Terra | OpenAI | 158.2 | — | 0.933 (max) | — | — | — |
| 7 | Claude Opus 4.8 | Anthropic | 158.0 | — | 0.910 (max) | — | — | — |
| 8 | GPT-5.4 Pro | OpenAI | 157.9 | — | 0.946 (xhigh) | — | — | — |
| 9 | GPT-5.4 | OpenAI | 156.1 | 0.769 (high) | 0.933 (xhigh) | — | — | — |
| 10 | Claude Opus 4.7 | Anthropic | 156.1 | 0.835 (max) | 0.902 (xhigh) | — | — | — |
| 11 | Kimi K3 | Moonshot | 155.6 | — | 0.931 (max) | — | — | — |
| 12 | GPT-5.3 Codex | OpenAI | 155.6 | 0.748 (high) | — | — | — | — |
| 13 | Claude Opus 4.6 | Anthropic | 155.4 | 0.772 | 0.905 (32k) | — | — | — |
| 14 | GPT-5.6 Luna | OpenAI | 155.3 | — | 0.916 (max) | — | — | — |
| 15 | Gemini 3.1 Pro Preview | Google DeepMind | 154.9 | 0.756 (customtools) | 0.941 | — | — | — |
| 16 | GPT-5.2 Pro | OpenAI | 154.7 | — | — | — | — | — |
| 17 | Gemini 3.5 Flash | 154.6 | 0.793 (high) | 0.928 (high) | — | — | — | |
| 18 | Muse Spark | Meta AI | 154.3 | — | 0.898 | — | — | — |
| 19 | GPT-5.2 | OpenAI | 153.7 | 0.738 (high) | 0.914 (xhigh) | — | — | — |
| 20 | Grok 4.5 | xAI | 153.6 | — | 0.934 (high) | — | — | — |
| 21 | Claude Sonnet 5 | Anthropic | 153.5 | — | 0.905 (xhigh) | — | — | — |
| 22 | Gemini 3 Pro Preview | Google DeepMind | 153.3 | 0.729 | 0.926 | — | — | — |
| 23 | Claude Sonnet 4.6 | Anthropic | 153.2 | 0.752 | 0.874 (32k) | — | — | — |
| 24 | Qwen3.7-Max | Alibaba | 153.1 | 0.773 | 0.916 | — | — | — |
| 25 | Grok 4.20 (4.20 0309 Reasoning) | xAI | 152.6 | — | 0.893 | — | — | — |
| 26 | Grok 4.20 | xAI | 152.6 | — | — | — | — | — |
| 27 | GLM-5.2 OPEN | Z.ai (Zhipu AI) | 151.4 | 0.787 (max) | 0.919 (max) | — | — | — |
| 28 | Kimi K2.6 OPEN | Moonshot | 151.0 | 0.767 | 0.908 | — | — | — |
| 29 | Gemini 3 Flash | Google DeepMind | 150.6 | 0.754 | 0.832 | — | — | — |
| 30 | GLM-5.1 | Z.ai (Zhipu AI) | 150.4 | 0.742 | 0.855 | — | — | — |
| 31 | GPT-5 Pro | OpenAI | 150.4 | — | — | — | — | — |
| 32 | GPT-5 (2025 08 07) | OpenAI | 150.0 | 0.736 (high) | 0.862 (high) | — | — | — |
| 33 | GPT-5 (Chat) | OpenAI | 150.0 | — | — | — | — | — |
| 34 | GPT-5.1 | OpenAI | 149.7 | 0.669 (high) | 0.876 (high) | — | — | — |
| 35 | Kimi K2.7 Code OPEN | Moonshot | 149.7 | — | 0.895 | — | — | — |
| 36 | Qwen 3.6 Max (Preview) | Alibaba | 149.7 | 0.767 | 0.891 | — | — | — |
| 37 | Claude Opus 4.5 | Anthropic | 149.6 | 0.767 | 0.860 (32k) | — | — | — |
| 38 | Qwen 3.6 Plus (2026-04-02) | Alibaba | 149.2 | 0.579 | 0.874 | — | — | — |
| 39 | Grok 4.3 Beta | xAI | 149.0 | — | 0.888 (high) | — | — | — |
| 40 | GPT-5.4 mini | OpenAI | 148.9 | — | 0.836 (high) | — | — | — |
| 41 | DeepSeek v4 Pro OPEN | DeepSeek | 148.7 | 0.776 (max) | 0.896 (max) | — | — | — |
| 42 | Kimi K2.5 (instant) OPEN | Moonshot | 148.3 | 0.738 | 0.876 | — | — | — |
| 43 | o3-pro | OpenAI | 147.9 | — | — | — | — | — |
| 44 | MiniMax-M2.5 OPEN | MiniMax | 147.1 | — | — | — | — | — |
| 45 | o3 | OpenAI | 147.1 | 0.623 (medium) | 0.818 (high) | — | — | — |
| 46 | Grok 4 | xAI | 147.0 | — | 0.870 | — | — | — |
| 47 | MiniMax-M3 | MiniMax | 147.0 | — | — | — | — | — |
| 48 | Qwen 3.5 Plus (hosted 397B-A17B) | Alibaba | 147.0 | — | 0.842 | — | — | — |
| 49 | Claude Sonnet 4.5 | Anthropic | 146.8 | 0.713 | 0.823 (59k) | — | — | — |
| 50 | GPT-5.4 nano | OpenAI | 146.6 | — | 0.785 (high) | — | — | — |
| 51 | DeepSeek-V3.2 (Thinking; Novita) OPEN | DeepSeek | 146.6 | — | — | — | — | — |
| 52 | DeepSeek-V3.2 (Reasoner) OPEN | DeepSeek | 146.6 | — | 0.834 | — | — | — |
| 53 | DeepSeek-V3.2 (Chat) OPEN | DeepSeek | 146.6 | — | — | — | — | — |
| 54 | o4-mini | OpenAI | 146.5 | — | 0.796 (high) | — | — | — |
| 55 | GLM-5 OPEN | Z.ai (Zhipu AI) | 146.4 | 0.721 | 0.878 | — | — | — |
| 56 | Gemini 2.5 Pro (Jun 2025) | Google DeepMind | 146.1 | 0.576 | 0.853 | — | — | — |
| 57 | Gemini 2.5 Pro Preview (Jun 2025) | Google DeepMind | 146.1 | — | 0.848 | — | — | — |
| 58 | Kimi K2 Thinking Turbo OPEN | Moonshot | 146.0 | — | 0.842 | — | — | — |
| 59 | Kimi K2 Thinking OPEN | Moonshot | 146.0 | — | — | — | — | — |
| 60 | GPT-5 mini | OpenAI | 145.8 | 0.647 (medium) | 0.750 (high) | — | — | — |
| 61 | MiniMax-M2.7 OPEN | MiniMax | 145.7 | — | — | — | — | — |
| 62 | Gemini 2.5 Pro Preview (Mar 2025) | Google DeepMind | 144.9 | — | — | — | — | — |
| 63 | Gemini 2.5 Pro Exp (Mar 2025) | Google DeepMind | 144.9 | — | 0.838 | — | — | — |
| 64 | DeepSeek-V3.2-Exp OPEN | DeepSeek | 144.9 | — | — | — | — | — |
| 65 | Qwen3-235B-A22B-Thinking-2507 OPEN | Alibaba | 144.8 | — | 0.801 | — | — | — |
| 66 | Grok 4 Fast | xAI | 144.6 | — | — | — | — | — |
| 67 | Gemini 3.1 Flash-Lite | 144.3 | — | — | — | — | — | |
| 68 | Claude Opus 4.1 | Anthropic | 144.2 | 0.733 | 0.773 (16k) | — | — | — |
| 69 | Qwen 3.6 Flash (2026-04-16) | Alibaba | 144.2 | — | 0.844 | — | — | — |
| 70 | Qwen 3.5 Flash (hosted 35B-A3B) | Alibaba | 144.1 | — | 0.838 | — | — | — |
| 71 | Qwen3-Max-Instruct | Alibaba | 143.8 | — | 0.726 | — | — | — |
| 72 | GLM-4.7 (Novita) OPEN | Z.ai (Zhipu AI) | 143.8 | — | 0.833 | — | — | — |
| 73 | Qwen 3.6 35B-A3B OPEN | Alibaba | 143.3 | — | — | — | — | — |
| 74 | Gemini 2.5 Flash (Preview 09 2025) | Google DeepMind | 143.2 | — | — | — | — | — |
| 75 | Claude Opus 4 | Anthropic | 142.9 | 0.707 | 0.763 (16k) | — | — | — |
| 76 | Claude Haiku 4.5 | Anthropic | 142.9 | — | 0.712 (32k) | — | — | — |
| 77 | Gemini 2.5 Pro Preview (May 2025) | Google DeepMind | 142.9 | — | 0.667 | — | — | — |
| 78 | o1 | OpenAI | 142.7 | — | 0.768 (high) | — | — | — |
| 79 | Gemini 2.5 Flash (Preview 05 20) | Google DeepMind | 142.2 | — | — | — | — | — |
| 80 | Claude Sonnet 4 | Anthropic | 142.2 | — | 0.783 (32k) | — | — | — |
| 81 | DeepSeek-R1 (May 2025) OPEN | DeepSeek | 142.0 | — | 0.763 | — | — | — |
| 82 | Gemma 4 31B IT OPEN | Google DeepMind | 141.9 | — | — | — | — | — |
| 83 | Claude 3.7 Sonnet | Anthropic | 141.7 | 0.610 | 0.785 (64k) | — | — | — |
| 84 | o3-mini | OpenAI | 141.4 | — | 0.770 (high) | — | — | — |
| 85 | Grok-3 mini | xAI | 141.0 | — | 0.763 (low) | — | — | — |
| 86 | Grok 3 mini | xAI | 141.0 | — | — | — | — | — |
| 87 | GLM-4.6 OPEN | Z.ai (Zhipu AI),Tsinghua University | 141.0 | — | — | — | — | — |
| 88 | GPT-5 nano | OpenAI | 140.9 | — | 0.694 (high) | — | — | — |
| 89 | Gemini 2.5 Flash (Preview 04 17) | Google DeepMind | 140.6 | — | — | — | — | — |
| 90 | gpt-oss-120b OPEN | OpenAI | 140.6 | — | 0.758 (high-openai) | — | — | — |
| 91 | Kimi K2 Instruct OPEN | Moonshot | 140.4 | — | — | — | — | — |
| 92 | Kimi K2 OPEN | Moonshot | 140.4 | — | — | — | — | — |
| 93 | Gemini 2.5 Flash | Google DeepMind | 140.3 | — | — | — | — | — |
| 94 | DeepSeek-R1 OPEN | DeepSeek | 139.7 | — | 0.692 | — | — | — |
| 95 | Qwen3-235B-A22B OPEN | Alibaba | 139.6 | — | 0.707 | — | — | — |
| 96 | Grok 3 (Beta) | xAI | 139.0 | — | 0.676 | — | — | — |
| 97 | Grok 3 | xAI | 139.0 | — | — | — | — | — |
| 98 | Qwen3-235B-A22B (Jul 2025) OPEN | Alibaba | 139.0 | — | — | — | — | — |
| 99 | Qwen3-235B-A22B (Sep 2025) OPEN | Alibaba | 139.0 | — | — | — | — | — |
| 100 | GPT-4.5 Preview (Feb 2025) | OpenAI | 137.6 | — | 0.687 | — | — | — |
| 101 | GPT-4.1 | OpenAI | 137.5 | 0.485 | 0.669 | — | — | — |
| 102 | DeepSeek-V3 (Mar 2025) OPEN | DeepSeek | 137.1 | — | 0.676 | — | — | — |
| 103 | o1-mini | OpenAI | 136.6 | — | 0.624 (high) | — | — | — |
| 104 | Gemini 2.0 Flash Thinking Exp | Google DeepMind,Google | 136.2 | — | 0.571 | — | — | — |
| 105 | o1-preview | OpenAI | 135.8 | — | 0.503 | — | — | — |
| 106 | GPT-4.1 mini | OpenAI | 135.8 | — | 0.658 | — | — | — |
| 107 | Gemini 2.0 Flash (Feb 2025) | Google DeepMind,Google | 135.8 | — | 0.641 | — | — | — |
| 108 | Gemini 2.0 Flash (2.0 Flash 02 05) | Google DeepMind,Google | 135.8 | — | — | — | — | — |
| 109 | Gemini 2.0 Pro Exp (Feb 2025) | Google DeepMind | 135.6 | — | 0.657 | — | — | — |
| 110 | Mistral Medium 3 | Mistral AI | 135.3 | — | 0.595 | — | — | — |
| 111 | Gemini 2.0 Flash (2.0 Flash Exp) | Google DeepMind,Google | 135.1 | — | — | — | — | — |
| 112 | Gemini 2.0 Flash (Exp 1206) | Google DeepMind,Google | 135.1 | — | — | — | — | — |
| 113 | Claude 3.5 Sonnet (Oct 2024) | Anthropic | 134.3 | — | 0.553 | — | — | — |
| 114 | Qwen2.5-Max (Qwen 2025 01 25) | Alibaba | 133.4 | — | 0.561 | — | — | — |
| 115 | Qwen2.5-Max | Alibaba | 133.4 | — | — | — | — | — |
| 116 | Magistral Small 1.1 OPEN | Mistral AI | 133.1 | — | 0.484 | — | — | — |
| 117 | Llama 4 Maverick (FP8) OPEN | Meta AI | 133.0 | — | 0.670 | — | — | — |
| 118 | Llama 4 Maverick OPEN | Meta AI | 133.0 | — | — | — | — | — |
| 119 | DeepSeek-V3 OPEN | DeepSeek | 133.0 | — | 0.565 | — | — | — |
| 120 | DeepSeek-V3 (Base) OPEN | DeepSeek | 133.0 | — | — | — | — | — |
| 121 | Gemini 1.5 Pro (002) | Google DeepMind | 132.7 | — | 0.572 | — | — | — |
| 122 | Phi-4 OPEN | Microsoft Research | 131.1 | — | 0.561 | — | — | — |
| 123 | Gemma 3 27B OPEN | Google DeepMind | 131.1 | — | 0.489 | — | — | — |
| 124 | GPT-4.1 nano | OpenAI | 130.9 | — | 0.489 | — | — | — |
| 125 | Grok-2 | xAI | 130.8 | — | 0.538 | — | — | — |
| 126 | Llama 4 Scout OPEN | Meta AI | 130.5 | — | 0.518 | — | — | — |
| 127 | Gemini 1.5 Flash (002) | Google DeepMind | 130.4 | — | 0.473 | — | — | — |
| 128 | Claude 3.5 Sonnet (Jun 2024) | Anthropic | 130.0 | — | 0.540 | — | — | — |
| 129 | GPT-4o (Nov 2024) | OpenAI | 129.3 | 0.310 | 0.479 | — | — | — |
| 130 | GPT-4o (Aug 2024) | OpenAI | 129.2 | — | 0.492 | — | — | — |
| 131 | Qwen2.5-72B (Instruct) OPEN | Alibaba | 129.1 | — | 0.491 | — | — | — |
| 132 | Qwen2.5-72B OPEN | Alibaba | 129.1 | — | — | — | — | — |
| 133 | Qwen2.5-72B (Vl Instruct) OPEN | Alibaba | 129.1 | — | — | — | — | — |
| 134 | Llama 3.1-405B (Instruct) OPEN | Meta AI | 129.0 | — | 0.509 | — | — | — |
| 135 | Llama 3.1-405B OPEN | Meta AI | 129.0 | — | — | — | — | — |
| 136 | GPT-4o (May 2024) | OpenAI | 128.8 | — | 0.489 | — | — | — |
| 137 | Mistral Large 2 (2411) OPEN | Mistral AI | 128.8 | — | 0.513 | — | — | — |
| 138 | Mistral Small 3.1 OPEN | Mistral AI | 127.7 | — | 0.475 | — | — | — |
| 139 | GPT-4 Turbo (Apr 2024) | OpenAI | 127.6 | — | 0.466 | — | — | — |
| 140 | Mistral Large 2 (2407) OPEN | Mistral AI | 127.5 | — | 0.490 | — | — | — |
| 141 | Claude 3.5 Haiku (Oct 2024) | Anthropic | 127.5 | — | 0.381 | — | — | — |
| 142 | Llama 3.3 70B OPEN | Meta AI | 127.5 | — | 0.474 | — | — | — |
| 143 | Gemini 1.5 Pro (001) | Google DeepMind | 127.2 | — | 0.459 | — | — | — |
| 144 | GPT-4o mini | OpenAI | 126.9 | — | 0.377 | — | — | — |
| 145 | Claude 3 Opus | Anthropic | 126.9 | — | 0.472 | — | — | — |
| 146 | GPT-4 (Mar 2023) | OpenAI | 125.9 | — | 0.357 | — | — | — |
| 147 | GPT-4 (Mar 2023) (32k) | OpenAI | 125.9 | — | — | — | — | — |
| 148 | Llama 3.2 90B OPEN | Meta AI | 125.7 | — | 0.410 | — | — | — |
| 149 | Llama 3.1-70B OPEN | Meta AI | 125.5 | — | 0.442 | — | — | — |
| 150 | Qwen2-72B OPEN | Alibaba | 125.5 | — | 0.408 | — | — | — |
| 151 | DeepSeek-V2 (MoE-236B) OPEN | DeepSeek | 124.6 | — | — | — | — | — |
| 152 | Amazon Nova Pro | Amazon | 123.9 | — | — | — | — | — |
| 153 | Llama 3-70B (Instruct) OPEN | Meta AI | 122.6 | — | 0.406 | — | — | — |
| 154 | Llama 3-70B OPEN | Meta AI | 122.6 | — | — | — | — | — |
| 155 | Gemini 1.5 Flash (May 2024) | Google DeepMind | 122.5 | — | 0.404 | — | — | — |
| 156 | Gemini 1.5 Flash (0514) | Google DeepMind | 122.5 | — | — | — | — | — |
| 157 | Gemma 2 27B OPEN | Google DeepMind | 122.5 | — | 0.365 | — | — | — |
| 158 | GPT-4 (Jun 2023) | OpenAI | 121.9 | — | 0.307 | — | — | — |
| 159 | phi-3-small 7.4B OPEN | Microsoft | 121.6 | — | — | — | — | — |
| 160 | Mixtral 8x22B (Open) OPEN | Mistral AI | 121.3 | — | 0.341 | — | — | — |
| 161 | Mixtral 8x22B (v0.1) OPEN | Mistral AI | 121.3 | — | — | — | — | — |
| 162 | Mixtral 8x22B (Instruct v0.1) OPEN | Mistral AI | 121.3 | — | — | — | — | — |
| 163 | Claude Instant (1.1) | Anthropic | 121.1 | — | — | — | — | — |
| 164 | Claude Instant (1.2) | Anthropic | 121.1 | — | — | — | — | — |
| 165 | phi-3-medium 14B OPEN | Microsoft | 121.0 | — | 0.276 | — | — | — |
| 166 | Mistral Large | Mistral AI | 121.0 | — | 0.388 | — | — | — |
| 167 | Claude 3 Sonnet | Anthropic | 119.9 | — | 0.406 | — | — | — |
| 168 | Claude 2 | Anthropic | 119.6 | — | 0.347 | — | — | — |
| 169 | Gemma 2 9B (It) OPEN | Google DeepMind | 119.5 | — | 0.275 | — | — | — |
| 170 | Gemma 2 9B OPEN | Google DeepMind | 119.5 | — | — | — | — | — |
| 171 | Qwen2.5-Coder (32B) OPEN | Alibaba | 119.2 | — | — | — | — | — |
| 172 | Mistral NeMo (Open) OPEN | Mistral AI | 118.4 | — | 0.299 | — | — | — |
| 173 | Mistral NeMo (Base) OPEN | Mistral AI | 118.4 | — | — | — | — | — |
| 174 | Mistral NeMo (Instruct) OPEN | Mistral AI | 118.4 | — | — | — | — | — |
| 175 | Claude 2.1 | Anthropic | 118.3 | — | 0.330 | — | — | — |
| 176 | GPT-3.5 Turbo (Nov 2023) | OpenAI | 118.2 | — | 0.280 | — | — | — |
| 177 | Mixtral 8x7B (Instruct v0.1) OPEN | Mistral AI | 118.1 | — | 0.306 | — | — | — |
| 178 | Mixtral 8x7B (Open) OPEN | Mistral AI | 118.1 | — | 0.298 | — | — | — |
| 179 | Mixtral 8x7B (v0.1) OPEN | Mistral AI | 118.1 | — | — | — | — | — |
| 180 | Claude 3 Haiku | Anthropic | 117.6 | — | 0.363 | — | — | — |
| 181 | Yi-34B (Chat) OPEN | 01.AI | 117.1 | — | 0.147 | — | — | — |
| 182 | Yi-34B OPEN | 01.AI | 117.1 | — | — | — | — | — |
| 183 | phi-3-mini 3.8B OPEN | Microsoft | 117.1 | — | — | — | — | — |
| 184 | Gemini 1.0 Pro | Google DeepMind | 117.0 | — | 0.340 | — | — | — |
| 185 | Stable Beluga 2 OPEN | Stability AI | 116.9 | — | — | — | — | — |
| 186 | Llama 3-8B (Instruct) OPEN | Meta AI | 116.4 | — | 0.261 | — | — | — |
| 187 | Llama 3-8B OPEN | Meta AI | 116.4 | — | — | — | — | — |
| 188 | Llama 3.1-8B OPEN | Meta AI | 115.4 | — | 0.259 | — | — | — |
| 189 | GPT-3.5 Turbo (Jan 2024) | OpenAI | 114.4 | — | 0.272 | — | — | — |
| 190 | Llama 2-70B (Chat Hf) OPEN | Meta AI | 113.7 | — | 0.263 | — | — | — |
| 191 | Llama 2-70B (Hf) OPEN | Meta AI | 113.7 | — | — | — | — | — |
| 192 | Llama 2-70B (Chat) OPEN | Meta AI | 113.7 | — | — | — | — | — |
| 193 | GPT-3.5 Turbo (June 2023) | OpenAI | 113.1 | — | — | — | — | — |
| 194 | Qwen2.5-Coder (7B) OPEN | Alibaba | 112.9 | — | — | — | — | — |
| 195 | Qwen2.5-Coder (7B) (Instruct) OPEN | Alibaba | 112.9 | — | — | — | — | — |
| 196 | Qwen-14B OPEN | Alibaba | 112.8 | — | — | — | — | — |
| 197 | Qwen-14B (Chat) OPEN | Alibaba | 112.8 | — | — | — | — | — |
| 198 | Mistral 7B OPEN | Mistral AI | 111.9 | — | — | — | — | — |
| 199 | Mistral 7B (Instruct) OPEN | Mistral AI | 111.9 | — | — | — | — | — |
| 200 | Falcon-180B OPEN | Technology Innovation Institute | 111.9 | — | — | — | — | — |
| 201 | Gemma 7B OPEN | Google DeepMind | 111.7 | — | — | — | — | — |
| 202 | LLaMA-65B OPEN | Meta AI | 109.9 | — | — | — | — | — |
| 203 | Falcon 2-11B OPEN | Technology Innovation Institute | 109.3 | — | — | — | — | — |
| 204 | Phi-2 OPEN | Microsoft | 107.6 | — | — | — | — | — |
| 205 | Nemotron-4 15B | NVIDIA | 107.4 | — | — | — | — | — |
| 206 | LLaMA-33B OPEN | Meta AI | 107.1 | — | — | — | — | — |
| 207 | Qwen-7B OPEN | Alibaba | 106.5 | — | — | — | — | — |
| 208 | Llama 2-13B OPEN | Meta AI | 105.9 | — | — | — | — | — |
| 209 | Llama 2-13B (Chat) OPEN | Meta AI | 105.9 | — | — | — | — | — |
| 210 | Llama 2-34B | Meta AI | 104.9 | — | — | — | — | — |
| 211 | StarCoder 2 15B OPEN | Hugging Face,ServiceNow,NVIDIA,BigCode | 104.7 | — | — | — | — | — |
| 212 | Yi 6B OPEN | 01.AI | 104.5 | — | — | — | — | — |
| 213 | Yi 6B (Chat) OPEN | 01.AI | 104.5 | — | — | — | — | — |
| 214 | Falcon-40B OPEN | Technology Innovation Institute | 104.1 | — | — | — | — | — |
| 215 | Falcon-40B (Instruct) OPEN | Technology Innovation Institute | 104.1 | — | — | — | — | — |
| 216 | Baichuan2-13B OPEN | Baichuan | 102.9 | — | — | — | — | — |
| 217 | Qwen2.5-Coder (1.5B) OPEN | Alibaba | 102.7 | — | — | — | — | — |
| 218 | INTELLECT-1 | Prime Intellect,Hugging Face,Arcee AI | 100.5 | — | — | — | — | — |
| 219 | MPT-30B OPEN | MosaicML | 100.3 | — | — | — | — | — |
| 220 | MPT-30B (Instruct) OPEN | MosaicML | 100.3 | — | — | — | — | — |
| 221 | LLaMA-13B OPEN | Meta AI | 100.2 | — | — | — | — | — |
| 222 | Llama 2-7B OPEN | Meta AI | 98.7 | — | — | — | — | — |
| 223 | LLaMA-7B OPEN | Meta AI | 96.3 | — | — | — | — | — |
| 224 | Baichuan 2-7B OPEN | Baichuan | 96.0 | — | — | — | — | — |
| 225 | DeepSeek Coder 33B OPEN | DeepSeek,Peking University | 96.0 | — | — | — | — | — |
| 226 | Falcon-7B OPEN | Technology Innovation Institute | 94.7 | — | — | — | — | — |
| 227 | MPT-7B OPEN | MosaicML | 94.2 | — | — | — | — | — |
| 228 | Gemma 2B OPEN | Google DeepMind | 93.8 | — | — | — | — | — |
| 229 | StarCoder 2 7B OPEN | Hugging Face,ServiceNow,NVIDIA,BigCode | 93.2 | — | — | — | — | — |
| 230 | XGen-7B OPEN | Salesforce | 93.0 | — | — | — | — | — |
| 231 | Phi-1.5 OPEN | Microsoft | 91.2 | — | — | — | — | — |
| 232 | Baichuan1-7B OPEN | Baichuan | 90.0 | — | — | — | — | — |
| 233 | Dolly 2.0-12b OPEN | Databricks | 89.3 | — | — | — | — | — |
| 234 | DeepSeek Coder 6.7B OPEN | DeepSeek,Peking University | 89.2 | — | — | — | — | — |
| 235 | StarCoder 2 3B OPEN | Hugging Face,ServiceNow,NVIDIA,BigCode | 88.4 | — | — | — | — | — |
| 236 | Cerebras-GPT-13B OPEN | Cerebras Systems | 82.8 | — | — | — | — | — |
| 237 | DeepSeek Coder 1.3B OPEN | DeepSeek,Peking University | 63.8 | — | — | — | — | — |
| Rank | Model | Provider | Overall (ECI) | Coding | Reasoning | Context | Speed | Reference Price |
|---|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 4.7 | Anthropic | 156.1 | 0.835 (max) | 0.902 (xhigh) | — | — | — |
| 2 | GPT-5.5 | OpenAI | 158.5 | 0.806 (xhigh-pre-release) | 0.940 (xhigh-pre-release) | — | — | — |
| 3 | Gemini 3.5 Flash | 154.6 | 0.793 (high) | 0.928 (high) | — | — | — | |
| 4 | GLM-5.2 OPEN | Z.ai (Zhipu AI) | 151.4 | 0.787 (max) | 0.919 (max) | — | — | — |
| 5 | DeepSeek v4 Pro OPEN | DeepSeek | 148.7 | 0.776 (max) | 0.896 (max) | — | — | — |
| 6 | Qwen3.7-Max | Alibaba | 153.1 | 0.773 | 0.916 | — | — | — |
| 7 | Claude Opus 4.6 | Anthropic | 155.4 | 0.772 | 0.905 (32k) | — | — | — |
| 8 | GPT-5.4 | OpenAI | 156.1 | 0.769 (high) | 0.933 (xhigh) | — | — | — |
| 9 | Qwen 3.6 Max (Preview) | Alibaba | 149.7 | 0.767 | 0.891 | — | — | — |
| 10 | Kimi K2.6 OPEN | Moonshot | 151.0 | 0.767 | 0.908 | — | — | — |
| 11 | Claude Opus 4.5 | Anthropic | 149.6 | 0.767 | 0.860 (32k) | — | — | — |
| 12 | Gemini 3.1 Pro Preview | Google DeepMind | 154.9 | 0.756 (customtools) | 0.941 | — | — | — |
| 13 | Gemini 3 Flash | Google DeepMind | 150.6 | 0.754 | 0.832 | — | — | — |
| 14 | Claude Sonnet 4.6 | Anthropic | 153.2 | 0.752 | 0.874 (32k) | — | — | — |
| 15 | GPT-5.3 Codex | OpenAI | 155.6 | 0.748 (high) | — | — | — | — |
| 16 | GLM-5.1 | Z.ai (Zhipu AI) | 150.4 | 0.742 | 0.855 | — | — | — |
| 17 | Kimi K2.5 (instant) OPEN | Moonshot | 148.3 | 0.738 | 0.876 | — | — | — |
| 18 | GPT-5.2 | OpenAI | 153.7 | 0.738 (high) | 0.914 (xhigh) | — | — | — |
| 19 | GPT-5 (2025 08 07) | OpenAI | 150.0 | 0.736 (high) | 0.862 (high) | — | — | — |
| 20 | Claude Opus 4.1 | Anthropic | 144.2 | 0.733 | 0.773 (16k) | — | — | — |
| 21 | Gemini 3 Pro Preview | Google DeepMind | 153.3 | 0.729 | 0.926 | — | — | — |
| 22 | GLM-5 OPEN | Z.ai (Zhipu AI) | 146.4 | 0.721 | 0.878 | — | — | — |
| 23 | Claude Sonnet 4.5 | Anthropic | 146.8 | 0.713 | 0.823 (59k) | — | — | — |
| 24 | Claude Opus 4 | Anthropic | 142.9 | 0.707 | 0.763 (16k) | — | — | — |
| 25 | GPT-5.1 | OpenAI | 149.7 | 0.669 (high) | 0.876 (high) | — | — | — |
| 26 | GPT-5 mini | OpenAI | 145.8 | 0.647 (medium) | 0.750 (high) | — | — | — |
| 27 | o3 | OpenAI | 147.1 | 0.623 (medium) | 0.818 (high) | — | — | — |
| 28 | Claude 3.7 Sonnet | Anthropic | 141.7 | 0.610 | 0.785 (64k) | — | — | — |
| 29 | Qwen 3.6 Plus (2026-04-02) | Alibaba | 149.2 | 0.579 | 0.874 | — | — | — |
| 30 | Gemini 2.5 Pro (Jun 2025) | Google DeepMind | 146.1 | 0.576 | 0.853 | — | — | — |
| 31 | GPT-4.1 | OpenAI | 137.5 | 0.485 | 0.669 | — | — | — |
| 32 | GPT-4o (Nov 2024) | OpenAI | 129.3 | 0.310 | 0.479 | — | — | — |
| Rank | Model | Provider | Overall (ECI) | Coding | Reasoning | Context | Speed | Reference Price |
|---|---|---|---|---|---|---|---|---|
| 1 | GPT-5.4 Pro | OpenAI | 157.9 | — | 0.946 (xhigh) | — | — | — |
| 2 | Gemini 3.1 Pro Preview | Google DeepMind | 154.9 | 0.756 (customtools) | 0.941 | — | — | — |
| 3 | GPT-5.5 | OpenAI | 158.5 | 0.806 (xhigh-pre-release) | 0.940 (xhigh-pre-release) | — | — | — |
| 4 | GPT-5.5 Pro | OpenAI | 160.9 | — | 0.939 (xhigh-pre-release) | — | — | — |
| 5 | Claude Opus 5 | Anthropic | 159.4 | — | 0.939 (max) | — | — | — |
| 6 | GPT-5.6 Sol | OpenAI | 161.7 | — | 0.935 (max) | — | — | — |
| 7 | Grok 4.5 | xAI | 153.6 | — | 0.934 (high) | — | — | — |
| 8 | GPT-5.6 Terra | OpenAI | 158.2 | — | 0.933 (max) | — | — | — |
| 9 | GPT-5.4 | OpenAI | 156.1 | 0.769 (high) | 0.933 (xhigh) | — | — | — |
| 10 | Kimi K3 | Moonshot | 155.6 | — | 0.931 (max) | — | — | — |
| 11 | Gemini 3.5 Flash | 154.6 | 0.793 (high) | 0.928 (high) | — | — | — | |
| 12 | Gemini 3 Pro Preview | Google DeepMind | 153.3 | 0.729 | 0.926 | — | — | — |
| 13 | GLM-5.2 OPEN | Z.ai (Zhipu AI) | 151.4 | 0.787 (max) | 0.919 (max) | — | — | — |
| 14 | GPT-5.6 Luna | OpenAI | 155.3 | — | 0.916 (max) | — | — | — |
| 15 | Qwen3.7-Max | Alibaba | 153.1 | 0.773 | 0.916 | — | — | — |
| 16 | GPT-5.2 | OpenAI | 153.7 | 0.738 (high) | 0.914 (xhigh) | — | — | — |
| 17 | Claude Opus 4.8 | Anthropic | 158.0 | — | 0.910 (max) | — | — | — |
| 18 | Kimi K2.6 OPEN | Moonshot | 151.0 | 0.767 | 0.908 | — | — | — |
| 19 | Claude Sonnet 5 | Anthropic | 153.5 | — | 0.905 (xhigh) | — | — | — |
| 20 | Claude Opus 4.6 | Anthropic | 155.4 | 0.772 | 0.905 (32k) | — | — | — |
| 21 | Claude Opus 4.7 | Anthropic | 156.1 | 0.835 (max) | 0.902 (xhigh) | — | — | — |
| 22 | Muse Spark | Meta AI | 154.3 | — | 0.898 | — | — | — |
| 23 | DeepSeek v4 Pro OPEN | DeepSeek | 148.7 | 0.776 (max) | 0.896 (max) | — | — | — |
| 24 | Kimi K2.7 Code OPEN | Moonshot | 149.7 | — | 0.895 | — | — | — |
| 25 | Grok 4.20 (4.20 0309 Reasoning) | xAI | 152.6 | — | 0.893 | — | — | — |
| 26 | Qwen 3.6 Max (Preview) | Alibaba | 149.7 | 0.767 | 0.891 | — | — | — |
| 27 | Grok 4.3 Beta | xAI | 149.0 | — | 0.888 (high) | — | — | — |
| 28 | GLM-5 OPEN | Z.ai (Zhipu AI) | 146.4 | 0.721 | 0.878 | — | — | — |
| 29 | GPT-5.1 | OpenAI | 149.7 | 0.669 (high) | 0.876 (high) | — | — | — |
| 30 | Kimi K2.5 (instant) OPEN | Moonshot | 148.3 | 0.738 | 0.876 | — | — | — |
| 31 | Qwen 3.6 Plus (2026-04-02) | Alibaba | 149.2 | 0.579 | 0.874 | — | — | — |
| 32 | Claude Sonnet 4.6 | Anthropic | 153.2 | 0.752 | 0.874 (32k) | — | — | — |
| 33 | Grok 4 | xAI | 147.0 | — | 0.870 | — | — | — |
| 34 | GPT-5 (2025 08 07) | OpenAI | 150.0 | 0.736 (high) | 0.862 (high) | — | — | — |
| 35 | Claude Opus 4.5 | Anthropic | 149.6 | 0.767 | 0.860 (32k) | — | — | — |
| 36 | GLM-5.1 | Z.ai (Zhipu AI) | 150.4 | 0.742 | 0.855 | — | — | — |
| 37 | Gemini 2.5 Pro (Jun 2025) | Google DeepMind | 146.1 | 0.576 | 0.853 | — | — | — |
| 38 | Gemini 2.5 Pro Preview (Jun 2025) | Google DeepMind | 146.1 | — | 0.848 | — | — | — |
| 39 | Qwen 3.6 Flash (2026-04-16) | Alibaba | 144.2 | — | 0.844 | — | — | — |
| 40 | Kimi K2 Thinking Turbo OPEN | Moonshot | 146.0 | — | 0.842 | — | — | — |
| 41 | Qwen 3.5 Plus (hosted 397B-A17B) | Alibaba | 147.0 | — | 0.842 | — | — | — |
| 42 | Qwen 3.5 Flash (hosted 35B-A3B) | Alibaba | 144.1 | — | 0.838 | — | — | — |
| 43 | Gemini 2.5 Pro Exp (Mar 2025) | Google DeepMind | 144.9 | — | 0.838 | — | — | — |
| 44 | GPT-5.4 mini | OpenAI | 148.9 | — | 0.836 (high) | — | — | — |
| 45 | DeepSeek-V3.2 (Reasoner) OPEN | DeepSeek | 146.6 | — | 0.834 | — | — | — |
| 46 | GLM-4.7 (Novita) OPEN | Z.ai (Zhipu AI) | 143.8 | — | 0.833 | — | — | — |
| 47 | Gemini 3 Flash | Google DeepMind | 150.6 | 0.754 | 0.832 | — | — | — |
| 48 | Claude Sonnet 4.5 | Anthropic | 146.8 | 0.713 | 0.823 (59k) | — | — | — |
| 49 | o3 | OpenAI | 147.1 | 0.623 (medium) | 0.818 (high) | — | — | — |
| 50 | Qwen3-235B-A22B-Thinking-2507 OPEN | Alibaba | 144.8 | — | 0.801 | — | — | — |
| 51 | o4-mini | OpenAI | 146.5 | — | 0.796 (high) | — | — | — |
| 52 | Claude 3.7 Sonnet | Anthropic | 141.7 | 0.610 | 0.785 (64k) | — | — | — |
| 53 | GPT-5.4 nano | OpenAI | 146.6 | — | 0.785 (high) | — | — | — |
| 54 | Claude Sonnet 4 | Anthropic | 142.2 | — | 0.783 (32k) | — | — | — |
| 55 | Claude Opus 4.1 | Anthropic | 144.2 | 0.733 | 0.773 (16k) | — | — | — |
| 56 | o3-mini | OpenAI | 141.4 | — | 0.770 (high) | — | — | — |
| 57 | o1 | OpenAI | 142.7 | — | 0.768 (high) | — | — | — |
| 58 | DeepSeek-R1 (May 2025) OPEN | DeepSeek | 142.0 | — | 0.763 | — | — | — |
| 59 | Grok-3 mini | xAI | 141.0 | — | 0.763 (low) | — | — | — |
| 60 | Claude Opus 4 | Anthropic | 142.9 | 0.707 | 0.763 (16k) | — | — | — |
| 61 | gpt-oss-120b OPEN | OpenAI | 140.6 | — | 0.758 (high-openai) | — | — | — |
| 62 | GPT-5 mini | OpenAI | 145.8 | 0.647 (medium) | 0.750 (high) | — | — | — |
| 63 | Qwen3-Max-Instruct | Alibaba | 143.8 | — | 0.726 | — | — | — |
| 64 | Claude Haiku 4.5 | Anthropic | 142.9 | — | 0.712 (32k) | — | — | — |
| 65 | Qwen3-235B-A22B OPEN | Alibaba | 139.6 | — | 0.707 | — | — | — |
| 66 | GPT-5 nano | OpenAI | 140.9 | — | 0.694 (high) | — | — | — |
| 67 | DeepSeek-R1 OPEN | DeepSeek | 139.7 | — | 0.692 | — | — | — |
| 68 | GPT-4.5 Preview (Feb 2025) | OpenAI | 137.6 | — | 0.687 | — | — | — |
| 69 | DeepSeek-V3 (Mar 2025) OPEN | DeepSeek | 137.1 | — | 0.676 | — | — | — |
| 70 | Grok 3 (Beta) | xAI | 139.0 | — | 0.676 | — | — | — |
| 71 | Llama 4 Maverick (FP8) OPEN | Meta AI | 133.0 | — | 0.670 | — | — | — |
| 72 | GPT-4.1 | OpenAI | 137.5 | 0.485 | 0.669 | — | — | — |
| 73 | Gemini 2.5 Pro Preview (May 2025) | Google DeepMind | 142.9 | — | 0.667 | — | — | — |
| 74 | GPT-4.1 mini | OpenAI | 135.8 | — | 0.658 | — | — | — |
| 75 | Gemini 2.0 Pro Exp (Feb 2025) | Google DeepMind | 135.6 | — | 0.657 | — | — | — |
| 76 | QWQ-Plus | Alibaba | — | — | 0.654 | — | — | — |
| 77 | Gemini 2.0 Flash (Feb 2025) | Google DeepMind,Google | 135.8 | — | 0.641 | — | — | — |
| 78 | o1-mini | OpenAI | 136.6 | — | 0.624 (high) | — | — | — |
| 79 | Mistral Medium 3 | Mistral AI | 135.3 | — | 0.595 | — | — | — |
| 80 | Gemini 1.5 Pro (002) | Google DeepMind | 132.7 | — | 0.572 | — | — | — |
| 81 | Gemini 2.0 Flash Thinking Exp | Google DeepMind,Google | 136.2 | — | 0.571 | — | — | — |
| 82 | DeepSeek-V3 OPEN | DeepSeek | 133.0 | — | 0.565 | — | — | — |
| 83 | Qwen2.5-Max (Qwen 2025 01 25) | Alibaba | 133.4 | — | 0.561 | — | — | — |
| 84 | Phi-4 OPEN | Microsoft Research | 131.1 | — | 0.561 | — | — | — |
| 85 | DeepSeek-R1-Distill-Llama-70B OPEN | DeepSeek | — | — | 0.557 | — | — | — |
| 86 | Claude 3.5 Sonnet (Oct 2024) | Anthropic | 134.3 | — | 0.553 | — | — | — |
| 87 | Claude 3.5 Sonnet (Jun 2024) | Anthropic | 130.0 | — | 0.540 | — | — | — |
| 88 | Grok-2 | xAI | 130.8 | — | 0.538 | — | — | — |
| 89 | Llama 4 Scout OPEN | Meta AI | 130.5 | — | 0.518 | — | — | — |
| 90 | Mistral Large 2 (2411) OPEN | Mistral AI | 128.8 | — | 0.513 | — | — | — |
| 91 | Llama 3.1-405B (Instruct) OPEN | Meta AI | 129.0 | — | 0.509 | — | — | — |
| 92 | o1-preview | OpenAI | 135.8 | — | 0.503 | — | — | — |
| 93 | GPT-4o (Aug 2024) | OpenAI | 129.2 | — | 0.492 | — | — | — |
| 94 | Qwen2.5-72B (Instruct) OPEN | Alibaba | 129.1 | — | 0.491 | — | — | — |
| 95 | Mistral Large 2 (2407) OPEN | Mistral AI | 127.5 | — | 0.490 | — | — | — |
| 96 | GPT-4.1 nano | OpenAI | 130.9 | — | 0.489 | — | — | — |
| 97 | GPT-4o (May 2024) | OpenAI | 128.8 | — | 0.489 | — | — | — |
| 98 | Gemma 3 27B OPEN | Google DeepMind | 131.1 | — | 0.489 | — | — | — |
| 99 | Magistral Small 1.1 OPEN | Mistral AI | 133.1 | — | 0.484 | — | — | — |
| 100 | Qwen Plus (01 25) | Alibaba | — | — | 0.481 | — | — | — |
| 101 | GPT-4o (Nov 2024) | OpenAI | 129.3 | 0.310 | 0.479 | — | — | — |
| 102 | Mistral Small 3.1 OPEN | Mistral AI | 127.7 | — | 0.475 | — | — | — |
| 103 | Llama 3.3 70B OPEN | Meta AI | 127.5 | — | 0.474 | — | — | — |
| 104 | Gemini 1.5 Flash (002) | Google DeepMind | 130.4 | — | 0.473 | — | — | — |
| 105 | Claude 3 Opus | Anthropic | 126.9 | — | 0.472 | — | — | — |
| 106 | GPT-4 Turbo (Apr 2024) | OpenAI | 127.6 | — | 0.466 | — | — | — |
| 107 | Tulu 3 (Tülu 3) 70B OPEN | Allen Institute for AI,University of Washington | — | — | 0.463 | — | — | — |
| 108 | Qwen2.5-32B OPEN | Alibaba | — | — | 0.461 | — | — | — |
| 109 | Gemini 1.5 Pro (001) | Google DeepMind | 127.2 | — | 0.459 | — | — | — |
| 110 | Mistral Small 3 OPEN | Mistral AI | — | — | 0.453 | — | — | — |
| 111 | DeepSeek-R1-Distill-Qwen-14B OPEN | DeepSeek | — | — | 0.447 | — | — | — |
| 112 | Llama 3.1-70B OPEN | Meta AI | 125.5 | — | 0.442 | — | — | — |
| 113 | WizardLM-2 8x22B OPEN | Microsoft | — | — | 0.434 | — | — | — |
| 114 | GPT-4 Turbo Preview (Nov 2023) | OpenAI | — | — | 0.424 | — | — | — |
| 115 | GPT-4 Turbo Preview (January 2024) | OpenAI | — | — | 0.423 | — | — | — |
| 116 | Qwen-Turbo | Alibaba | — | — | 0.418 | — | — | — |
| 117 | Llama 3.2 90B OPEN | Meta AI | 125.7 | — | 0.410 | — | — | — |
| 118 | Qwen2-72B OPEN | Alibaba | 125.5 | — | 0.408 | — | — | — |
| 119 | Claude 3 Sonnet | Anthropic | 119.9 | — | 0.406 | — | — | — |
| 120 | Llama 3-70B (Instruct) OPEN | Meta AI | 122.6 | — | 0.406 | — | — | — |
| 121 | Gemini 1.5 Flash (May 2024) | Google DeepMind | 122.5 | — | 0.404 | — | — | — |
| 122 | Mistral Large | Mistral AI | 121.0 | — | 0.388 | — | — | — |
| 123 | Claude 3.5 Haiku (Oct 2024) | Anthropic | 127.5 | — | 0.381 | — | — | — |
| 124 | GPT-4o mini | OpenAI | 126.9 | — | 0.377 | — | — | — |
| 125 | Hermes 2 Theta Llama-3 70B OPEN | Nous Research,Arcee AI | — | — | 0.375 | — | — | — |
| 126 | Gemma 2 27B OPEN | Google DeepMind | 122.5 | — | 0.365 | — | — | — |
| 127 | Claude 3 Haiku | Anthropic | 117.6 | — | 0.363 | — | — | — |
| 128 | GPT-4 (Mar 2023) | OpenAI | 125.9 | — | 0.357 | — | — | — |
| 129 | Claude 2 | Anthropic | 119.6 | — | 0.347 | — | — | — |
| 130 | Mixtral 8x22B (Open) OPEN | Mistral AI | 121.3 | — | 0.341 | — | — | — |
| 131 | Gemini 1.0 Pro | Google DeepMind | 117.0 | — | 0.340 | — | — | — |
| 132 | Eurus-2-7B-PRIME OPEN | Tsinghua University,University of Illinois Urbana-Champaign (UIUC),Shanghai AI Lab,Peking University,Shanghai Jiao Tong University,CUHK Shenzhen Research Institute | — | — | 0.339 | — | — | — |
| 133 | Gemini 1.5 Flash 8B | Google DeepMind | — | — | 0.330 | — | — | — |
| 134 | Claude 2.1 | Anthropic | 118.3 | — | 0.330 | — | — | — |
| 135 | DBRX OPEN | Databricks | — | — | 0.329 | — | — | — |
| 136 | Yi-1.5-34B OPEN | 01.AI | — | — | 0.320 | — | — | — |
| 137 | Qwen1.5-32B (Chat) OPEN | Alibaba | — | — | 0.307 | — | — | — |
| 138 | GPT-4 (Jun 2023) | OpenAI | 121.9 | — | 0.307 | — | — | — |
| 139 | Mixtral 8x7B (Instruct v0.1) OPEN | Mistral AI | 118.1 | — | 0.306 | — | — | — |
| 140 | Mistral NeMo (Open) OPEN | Mistral AI | 118.4 | — | 0.299 | — | — | — |
| 141 | Mixtral 8x7B (Open) OPEN | Mistral AI | 118.1 | — | 0.298 | — | — | — |
| 142 | Qwen1.5-72B OPEN | Alibaba | — | — | 0.288 | — | — | — |
| 143 | GPT-3.5 Turbo (Nov 2023) | OpenAI | 118.2 | — | 0.280 | — | — | — |
| 144 | phi-3-medium 14B OPEN | Microsoft | 121.0 | — | 0.276 | — | — | — |
| 145 | Gemma 2 9B (It) OPEN | Google DeepMind | 119.5 | — | 0.275 | — | — | — |
| 146 | GPT-3.5 Turbo (Jan 2024) | OpenAI | 114.4 | — | 0.272 | — | — | — |
| 147 | Ministral 8B OPEN | Mistral AI | — | — | 0.271 | — | — | — |
| 148 | Llama 2-70B (Chat Hf) OPEN | Meta AI | 113.7 | — | 0.263 | — | — | — |
| 149 | Llama 3-8B (Instruct) OPEN | Meta AI | 116.4 | — | 0.261 | — | — | — |
| 150 | Llama 3.1-8B OPEN | Meta AI | 115.4 | — | 0.259 | — | — | — |
| 151 | Ministral 3B | Mistral AI | — | — | 0.253 | — | — | — |
| 152 | DeepSeek LLM 67B OPEN | DeepSeek | — | — | 0.246 | — | — | — |
| 153 | Mistral 7B (Instruct v0.3) OPEN | Mistral AI | — | — | 0.152 | — | — | — |
| 154 | Yi-34B (Chat) OPEN | 01.AI | 117.1 | — | 0.147 | — | — | — |
| 155 | Mistral 7B (Open) OPEN | Mistral AI | — | — | 0.132 | — | — | — |
No Epoch-run agentic benchmark is ingested yet. Epoch redistributes several agentic third-party leaderboards (apex_agents_external, the_agent_company_external, os_world_external and others) inside the same bundle, but each needs its own licensing review before Xentir reads it — see Data Sources. This tab will populate once one clears.
| Rank | Model | Provider | Overall (ECI) | Coding | Reasoning | Context | Speed | Reference Price |
|---|---|---|---|---|---|---|---|---|
| 1 | GLM-5.2 OPEN | Z.ai (Zhipu AI) | 151.4 | 0.787 (max) | 0.919 (max) | — | — | — |
| 2 | Kimi K2.6 OPEN | Moonshot | 151.0 | 0.767 | 0.908 | — | — | — |
| 3 | Kimi K2.7 Code OPEN | Moonshot | 149.7 | — | 0.895 | — | — | — |
| 4 | DeepSeek v4 Pro OPEN | DeepSeek | 148.7 | 0.776 (max) | 0.896 (max) | — | — | — |
| 5 | Kimi K2.5 (instant) OPEN | Moonshot | 148.3 | 0.738 | 0.876 | — | — | — |
| 6 | MiniMax-M2.5 OPEN | MiniMax | 147.1 | — | — | — | — | — |
| 7 | DeepSeek-V3.2 (Thinking; Novita) OPEN | DeepSeek | 146.6 | — | — | — | — | — |
| 8 | DeepSeek-V3.2 (Reasoner) OPEN | DeepSeek | 146.6 | — | 0.834 | — | — | — |
| 9 | DeepSeek-V3.2 (Chat) OPEN | DeepSeek | 146.6 | — | — | — | — | — |
| 10 | GLM-5 OPEN | Z.ai (Zhipu AI) | 146.4 | 0.721 | 0.878 | — | — | — |
| 11 | Kimi K2 Thinking Turbo OPEN | Moonshot | 146.0 | — | 0.842 | — | — | — |
| 12 | Kimi K2 Thinking OPEN | Moonshot | 146.0 | — | — | — | — | — |
| 13 | MiniMax-M2.7 OPEN | MiniMax | 145.7 | — | — | — | — | — |
| 14 | DeepSeek-V3.2-Exp OPEN | DeepSeek | 144.9 | — | — | — | — | — |
| 15 | Qwen3-235B-A22B-Thinking-2507 OPEN | Alibaba | 144.8 | — | 0.801 | — | — | — |
| 16 | GLM-4.7 (Novita) OPEN | Z.ai (Zhipu AI) | 143.8 | — | 0.833 | — | — | — |
| 17 | Qwen 3.6 35B-A3B OPEN | Alibaba | 143.3 | — | — | — | — | — |
| 18 | DeepSeek-R1 (May 2025) OPEN | DeepSeek | 142.0 | — | 0.763 | — | — | — |
| 19 | Gemma 4 31B IT OPEN | Google DeepMind | 141.9 | — | — | — | — | — |
| 20 | GLM-4.6 OPEN | Z.ai (Zhipu AI),Tsinghua University | 141.0 | — | — | — | — | — |
| 21 | gpt-oss-120b OPEN | OpenAI | 140.6 | — | 0.758 (high-openai) | — | — | — |
| 22 | Kimi K2 Instruct OPEN | Moonshot | 140.4 | — | — | — | — | — |
| 23 | Kimi K2 OPEN | Moonshot | 140.4 | — | — | — | — | — |
| 24 | DeepSeek-R1 OPEN | DeepSeek | 139.7 | — | 0.692 | — | — | — |
| 25 | Qwen3-235B-A22B OPEN | Alibaba | 139.6 | — | 0.707 | — | — | — |
| 26 | Qwen3-235B-A22B (Jul 2025) OPEN | Alibaba | 139.0 | — | — | — | — | — |
| 27 | Qwen3-235B-A22B (Sep 2025) OPEN | Alibaba | 139.0 | — | — | — | — | — |
| 28 | DeepSeek-V3 (Mar 2025) OPEN | DeepSeek | 137.1 | — | 0.676 | — | — | — |
| 29 | Magistral Small 1.1 OPEN | Mistral AI | 133.1 | — | 0.484 | — | — | — |
| 30 | Llama 4 Maverick (FP8) OPEN | Meta AI | 133.0 | — | 0.670 | — | — | — |
| 31 | Llama 4 Maverick OPEN | Meta AI | 133.0 | — | — | — | — | — |
| 32 | DeepSeek-V3 OPEN | DeepSeek | 133.0 | — | 0.565 | — | — | — |
| 33 | DeepSeek-V3 (Base) OPEN | DeepSeek | 133.0 | — | — | — | — | — |
| 34 | Phi-4 OPEN | Microsoft Research | 131.1 | — | 0.561 | — | — | — |
| 35 | Gemma 3 27B OPEN | Google DeepMind | 131.1 | — | 0.489 | — | — | — |
| 36 | Llama 4 Scout OPEN | Meta AI | 130.5 | — | 0.518 | — | — | — |
| 37 | Qwen2.5-72B (Instruct) OPEN | Alibaba | 129.1 | — | 0.491 | — | — | — |
| 38 | Qwen2.5-72B OPEN | Alibaba | 129.1 | — | — | — | — | — |
| 39 | Qwen2.5-72B (Vl Instruct) OPEN | Alibaba | 129.1 | — | — | — | — | — |
| 40 | Llama 3.1-405B (Instruct) OPEN | Meta AI | 129.0 | — | 0.509 | — | — | — |
| 41 | Llama 3.1-405B OPEN | Meta AI | 129.0 | — | — | — | — | — |
| 42 | Mistral Large 2 (2411) OPEN | Mistral AI | 128.8 | — | 0.513 | — | — | — |
| 43 | Mistral Small 3.1 OPEN | Mistral AI | 127.7 | — | 0.475 | — | — | — |
| 44 | Mistral Large 2 (2407) OPEN | Mistral AI | 127.5 | — | 0.490 | — | — | — |
| 45 | Llama 3.3 70B OPEN | Meta AI | 127.5 | — | 0.474 | — | — | — |
| 46 | Llama 3.2 90B OPEN | Meta AI | 125.7 | — | 0.410 | — | — | — |
| 47 | Llama 3.1-70B OPEN | Meta AI | 125.5 | — | 0.442 | — | — | — |
| 48 | Qwen2-72B OPEN | Alibaba | 125.5 | — | 0.408 | — | — | — |
| 49 | DeepSeek-V2 (MoE-236B) OPEN | DeepSeek | 124.6 | — | — | — | — | — |
| 50 | Llama 3-70B (Instruct) OPEN | Meta AI | 122.6 | — | 0.406 | — | — | — |
| 51 | Llama 3-70B OPEN | Meta AI | 122.6 | — | — | — | — | — |
| 52 | Gemma 2 27B OPEN | Google DeepMind | 122.5 | — | 0.365 | — | — | — |
| 53 | phi-3-small 7.4B OPEN | Microsoft | 121.6 | — | — | — | — | — |
| 54 | Mixtral 8x22B (Open) OPEN | Mistral AI | 121.3 | — | 0.341 | — | — | — |
| 55 | Mixtral 8x22B (v0.1) OPEN | Mistral AI | 121.3 | — | — | — | — | — |
| 56 | Mixtral 8x22B (Instruct v0.1) OPEN | Mistral AI | 121.3 | — | — | — | — | — |
| 57 | phi-3-medium 14B OPEN | Microsoft | 121.0 | — | 0.276 | — | — | — |
| 58 | Gemma 2 9B (It) OPEN | Google DeepMind | 119.5 | — | 0.275 | — | — | — |
| 59 | Gemma 2 9B OPEN | Google DeepMind | 119.5 | — | — | — | — | — |
| 60 | Qwen2.5-Coder (32B) OPEN | Alibaba | 119.2 | — | — | — | — | — |
| 61 | Mistral NeMo (Open) OPEN | Mistral AI | 118.4 | — | 0.299 | — | — | — |
| 62 | Mistral NeMo (Base) OPEN | Mistral AI | 118.4 | — | — | — | — | — |
| 63 | Mistral NeMo (Instruct) OPEN | Mistral AI | 118.4 | — | — | — | — | — |
| 64 | Mixtral 8x7B (Instruct v0.1) OPEN | Mistral AI | 118.1 | — | 0.306 | — | — | — |
| 65 | Mixtral 8x7B (Open) OPEN | Mistral AI | 118.1 | — | 0.298 | — | — | — |
| 66 | Mixtral 8x7B (v0.1) OPEN | Mistral AI | 118.1 | — | — | — | — | — |
| 67 | Yi-34B (Chat) OPEN | 01.AI | 117.1 | — | 0.147 | — | — | — |
| 68 | Yi-34B OPEN | 01.AI | 117.1 | — | — | — | — | — |
| 69 | phi-3-mini 3.8B OPEN | Microsoft | 117.1 | — | — | — | — | — |
| 70 | Stable Beluga 2 OPEN | Stability AI | 116.9 | — | — | — | — | — |
| 71 | Llama 3-8B (Instruct) OPEN | Meta AI | 116.4 | — | 0.261 | — | — | — |
| 72 | Llama 3-8B OPEN | Meta AI | 116.4 | — | — | — | — | — |
| 73 | Llama 3.1-8B OPEN | Meta AI | 115.4 | — | 0.259 | — | — | — |
| 74 | Llama 2-70B (Chat Hf) OPEN | Meta AI | 113.7 | — | 0.263 | — | — | — |
| 75 | Llama 2-70B (Hf) OPEN | Meta AI | 113.7 | — | — | — | — | — |
| 76 | Llama 2-70B (Chat) OPEN | Meta AI | 113.7 | — | — | — | — | — |
| 77 | Qwen2.5-Coder (7B) OPEN | Alibaba | 112.9 | — | — | — | — | — |
| 78 | Qwen2.5-Coder (7B) (Instruct) OPEN | Alibaba | 112.9 | — | — | — | — | — |
| 79 | Qwen-14B OPEN | Alibaba | 112.8 | — | — | — | — | — |
| 80 | Qwen-14B (Chat) OPEN | Alibaba | 112.8 | — | — | — | — | — |
| 81 | Mistral 7B OPEN | Mistral AI | 111.9 | — | — | — | — | — |
| 82 | Mistral 7B (Instruct) OPEN | Mistral AI | 111.9 | — | — | — | — | — |
| 83 | Falcon-180B OPEN | Technology Innovation Institute | 111.9 | — | — | — | — | — |
| 84 | Gemma 7B OPEN | Google DeepMind | 111.7 | — | — | — | — | — |
| 85 | LLaMA-65B OPEN | Meta AI | 109.9 | — | — | — | — | — |
| 86 | Falcon 2-11B OPEN | Technology Innovation Institute | 109.3 | — | — | — | — | — |
| 87 | Phi-2 OPEN | Microsoft | 107.6 | — | — | — | — | — |
| 88 | LLaMA-33B OPEN | Meta AI | 107.1 | — | — | — | — | — |
| 89 | Qwen-7B OPEN | Alibaba | 106.5 | — | — | — | — | — |
| 90 | Llama 2-13B OPEN | Meta AI | 105.9 | — | — | — | — | — |
| 91 | Llama 2-13B (Chat) OPEN | Meta AI | 105.9 | — | — | — | — | — |
| 92 | StarCoder 2 15B OPEN | Hugging Face,ServiceNow,NVIDIA,BigCode | 104.7 | — | — | — | — | — |
| 93 | Yi 6B OPEN | 01.AI | 104.5 | — | — | — | — | — |
| 94 | Yi 6B (Chat) OPEN | 01.AI | 104.5 | — | — | — | — | — |
| 95 | Falcon-40B OPEN | Technology Innovation Institute | 104.1 | — | — | — | — | — |
| 96 | Falcon-40B (Instruct) OPEN | Technology Innovation Institute | 104.1 | — | — | — | — | — |
| 97 | Baichuan2-13B OPEN | Baichuan | 102.9 | — | — | — | — | — |
| 98 | Qwen2.5-Coder (1.5B) OPEN | Alibaba | 102.7 | — | — | — | — | — |
| 99 | MPT-30B OPEN | MosaicML | 100.3 | — | — | — | — | — |
| 100 | MPT-30B (Instruct) OPEN | MosaicML | 100.3 | — | — | — | — | — |
| 101 | LLaMA-13B OPEN | Meta AI | 100.2 | — | — | — | — | — |
| 102 | Llama 2-7B OPEN | Meta AI | 98.7 | — | — | — | — | — |
| 103 | LLaMA-7B OPEN | Meta AI | 96.3 | — | — | — | — | — |
| 104 | Baichuan 2-7B OPEN | Baichuan | 96.0 | — | — | — | — | — |
| 105 | DeepSeek Coder 33B OPEN | DeepSeek,Peking University | 96.0 | — | — | — | — | — |
| 106 | Falcon-7B OPEN | Technology Innovation Institute | 94.7 | — | — | — | — | — |
| 107 | MPT-7B OPEN | MosaicML | 94.2 | — | — | — | — | — |
| 108 | Gemma 2B OPEN | Google DeepMind | 93.8 | — | — | — | — | — |
| 109 | StarCoder 2 7B OPEN | Hugging Face,ServiceNow,NVIDIA,BigCode | 93.2 | — | — | — | — | — |
| 110 | XGen-7B OPEN | Salesforce | 93.0 | — | — | — | — | — |
| 111 | Phi-1.5 OPEN | Microsoft | 91.2 | — | — | — | — | — |
| 112 | Baichuan1-7B OPEN | Baichuan | 90.0 | — | — | — | — | — |
| 113 | Dolly 2.0-12b OPEN | Databricks | 89.3 | — | — | — | — | — |
| 114 | DeepSeek Coder 6.7B OPEN | DeepSeek,Peking University | 89.2 | — | — | — | — | — |
| 115 | StarCoder 2 3B OPEN | Hugging Face,ServiceNow,NVIDIA,BigCode | 88.4 | — | — | — | — | — |
| 116 | Cerebras-GPT-13B OPEN | Cerebras Systems | 82.8 | — | — | — | — | — |
| 117 | DeepSeek Coder 1.3B OPEN | DeepSeek,Peking University | 63.8 | — | — | — | — | — |
Xentir