AI Model Leaderboard
As of 2026-10-18, the highest-accuracy AI model on the un-leaked meo benchmark is OpenAI: GPT-5.5 (94.1%), while Owl Alpha offers the best intelligence-per-dollar at $0.0000 per correct answer. The board fuses accuracy, speed, cost, and error-cascade into Effective Value (π) β and the ranking inverts with task depth: fast models win one-shot tasks, accurate models dominate deep agentic chains. Data syncs hourly from the meo v2 API; new models typically appear within hours of release.
Movers: Β· all movers β
Columns
Effective Value (π) β re-rank by task depth
| Lab | Weights | Arena Elo | AA Intel | GPQA | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | OpenAI: GPT-5.5 | openai | 94.1% | 23.2 | 0.010 | $0.0336 | 22 | 1,067 | 1.1M | $5.00 | $30.00 | β | β | 60.2 | 94% |
| 2 | Google: Gemini 4 Argon | 92.5% | 18.2 | β | $0.0246 | 33 | 2,305 | 1.0M | $2.00 | $10.00 | β | β | β | β | |
| 3 | Anthropic: Claude Opus 5.5 | anthropic | 92.1% | 100.0 | β | $0.0165 | 7.72 | 734 | 1M | $4.00 | $20.00 | β | β | β | β |
| 4 | OpenAI: GPT-5.5 Pro | openai | 91.6% | 4.3 | β | $0.3197 | 56 | 1,722 | 1.1M | $30.00 | $180.00 | β | β | β | β |
| 5 | OpenAI: GPT-6.1 Sol | openai | 90.4% | 35.6 | β | $0.0036 | 7.34 | 297 | 1.1M | $2.00 | $10.00 | β | β | β | β |
| 6 | Anthropic: Claude Fable 5 | anthropic | 88.7% | 27.5 | β | $0.0573 | 14 | 1,053 | 1M | $10.00 | $50.00 | β | β | β | β |
| 7 | Anthropic: Claude Opus 4.8 | anthropic | 84.9% | 12.1 | -0.031 | $0.0459 | 21 | 1,740 | 1M | $5.00 | $25.00 | β | β | 61.4 | 92% |
| 8 | inclusionAI: Ring-2.6-1T | inclusionai | 84.7% | 14.8 | β | $0.0024 | 27 | 3,763 | 262K | $0.07 | $0.63 | β | β | 38.5 | 86% |
| 9 | Z.ai: GLM 5.2 | z-ai | 83.2% | 0.78 | β | $0.0200 | 110 | 4,159 | 1.0M | $0.04 | $7.00 | Open | β | β | β |
| 10 | Google: Gemini 3.5 Flash | 82.0% | 11.1 | β | $0.0432 | 28 | 4,666 | 1.0M | $1.50 | $9.00 | β | β | 55.3 | 92% | |
| 11 | DeepSeek: DeepSeek V4 Flash 0423 | deepseek | 81.8% | 0.43 | β | $0.0016 | 169 | 6,544 | 1.0M | $0.09 | $0.18 | Open | β | 46.5 | 89% |
| 12 | Qwen: Qwen3.7 Max | qwen | 81.0% | 0.65 | β | $0.0157 | 102 | 4,052 | 1M | $1.48 | $4.42 | β | β | 56.6 | 92% |
| 13 | MoonshotAI: Kimi K3 | moonshotai | 80.3% | 3.6 | β | $0.0379 | 32 | 2,532 | 1.0M | $0.45 | $15.00 | Open | β | β | β |
| 14 | DeepSeek: DeepSeek V4 Pro 0423 | deepseek | 80.3% | 0.53 | β | $0.0182 | 131 | 6,043 | 1.0M | $0.96 | $1.91 | Open | β | 51.5 | 89% |
| 15 | Google: Gemini 3.1 Pro Preview | 79.5% | 3.2 | 0.021 | $0.0597 | 44 | 4,837 | 1.0M | $2.00 | $12.00 | β | β | 57.2 | 94% | |
| 16 | NVIDIA: Nemotron 3 Ultra | nvidia | 75.2% | 1.2 | β | $0.0224 | 72 | 8,875 | 262K | $0.60 | $2.40 | Open | β | β | β |
| 17 | Owl Alpha | openrouter | 73.7% | 0.37 | β | $0.0000 | 53 | 1,982 | 1.0M | $0.00 | $0.00 | β | β | β | β |
| 18 | SpaceXAI: Grok 4.7 | x-ai | 72.4% | 0.17 | β | $0.0579 | 151 | 9,335 | 500K | $2.00 | $6.00 | β | β | β | β |
| 19 | Xiaomi: MiMo-V2.5-Pro | xiaomi | 71.5% | 0.18 | β | $0.0244 | 129 | 8,004 | 1.1M | $0.43 | $0.87 | Open | β | 53.8 | 87% |
| 20 | OpenAI: GPT-5.4 Nano | openai | 69.5% | 2.3 | β | $0.0029 | 16 | 2,265 | 400K | $0.20 | $1.25 | β | β | 44.0 | 82% |
| 21 | Qwen: Qwen3.6 Flash | qwen | 69.5% | 0.17 | β | $0.0157 | 147 | 13,896 | 1M | $0.19 | $1.13 | β | β | β | β |
| 22 | Qwen: Qwen3.5-Flash | qwen | 68.2% | 0.15 | β | $0.0051 | 165 | 19,567 | 1M | $0.07 | $0.26 | β | β | β | β |
| 23 | MiniMax: MiniMax M3 | minimax | 67.4% | 0.13 | β | $0.0047 | 73 | 3,794 | 1.0M | $0.30 | $1.20 | β | β | 54.7 | 93% |
| 24 | MoonshotAI: Kimi K2.6 | moonshotai | 66.9% | 0.050 | β | $0.0281 | 159 | 7,485 | 262K | $0.65 | $3.41 | Open | β | β | β |
| 25 | MoonshotAI: Kimi K2.7 Code | moonshotai | 66.9% | 0.073 | β | $0.0139 | 88 | 3,392 | 262K | $0.67 | $3.35 | Open | β | β | β |
| 26 | OpenAI: GPT-5 Mini | openai | 66.9% | 0.25 | β | $0.0102 | 58 | 5,039 | 400K | $0.25 | $2.00 | β | β | 41.2 | 83% |
| 27 | Qwen: Qwen3.7 Plus | qwen | 66.9% | 0.051 | β | $0.0083 | 128 | 5,065 | 1M | $0.32 | $1.28 | β | β | 53.3 | 90% |
| 28 | OpenAI: GPT-5.1-Codex-Mini | openai | 66.5% | 1.3 | β | $0.0082 | 22 | 4,029 | 400K | $0.25 | $2.00 | β | β | 38.6 | 81% |
| 29 | SpaceXAI: Grok 4.3 | x-ai | 66.5% | 1.7 | -0.031 | $0.0127 | 20 | 4,004 | 1M | $1.25 | $2.50 | β | β | 53.2 | 90% |
| 30 | Xiaomi: MiMo-V2.5 | xiaomi | 66.5% | 0.26 | β | $0.0016 | 57 | 5,332 | 1.1M | $0.14 | $0.28 | Open | β | β | β |
| 31 | Qwen: Qwen3 VL 30B A3B Instruct | qwen | 66.1% | 0.13 | β | $0.0023 | 64 | 3,599 | 262K | $0.15 | $0.60 | Open | β | β | β |
| 32 | Qwen: Qwen3 VL 32B Instruct | qwen | 66.1% | 0.19 | β | $0.0011 | 45 | 2,625 | 262K | $0.10 | $0.42 | Open | β | β | β |
| 33 | Nex AGI: Nex-N2-Pro (free) | nex-agi | 65.7% | 0.18 | β | $0.0000 | 37 | 1,913 | 262K | $0.00 | $0.00 | Open | β | β | β |
| 34 | Z.ai: GLM 4.6V | z-ai | 65.7% | 0.12 | β | $0.0035 | 64 | 3,740 | 131K | $0.30 | $0.90 | Open | β | β | β |
| 35 | ByteDance Seed: Seed 1.6 | bytedance-seed | 64.9% | 0.036 | β | $0.0069 | 104 | 3,384 | 262K | $0.25 | $2.00 | β | β | β | β |
| 36 | Qwen: Qwen3 VL 235B A22B Instruct | qwen | 64.4% | 0.060 | β | $0.0028 | 61 | 2,099 | 262K | $0.21 | $1.90 | Open | β | β | β |
| 37 | Qwen: Qwen3 VL 30B A3B Thinking | qwen | 64.0% | 0.008 | β | $0.0068 | 278 | 6,617 | 262K | $0.29 | $1.00 | Open | β | β | β |
| 38 | Google: Gemma 4 31B | 63.6% | 0.009 | β | $0.0030 | 266 | 6,611 | 262K | $0.20 | $0.40 | Open | β | 39.2 | 86% | |
| 39 | OpenAI: GPT-4.1 Mini | openai | 63.6% | 0.26 | β | $0.0023 | 22 | 1,312 | 1.0M | $0.40 | $1.60 | β | 1337 | 22.9 | 66% |
| 40 | Z.ai: GLM 5.1 | z-ai | 63.5% | 0.033 | β | $0.0426 | 170 | 10,581 | 205K | $0.97 | $3.04 | Open | β | 51.4 | 87% |
| 41 | Baidu: ERNIE 4.5 VL 424B A47B | baidu | 63.2% | 0.015 | β | $0.0087 | 192 | 6,546 | 131K | $0.42 | $1.25 | Open | β | β | β |
| 43 | Qwen: Qwen3 VL 8B Thinking | qwen | 63.2% | 0.072 | β | $0.0138 | 111 | 10,076 | 256K | $0.12 | $1.36 | Open | β | β | β |
| 44 | Google: Gemini 3.1 Flash Lite Preview | 61.9% | 0.46 | β | $0.0128 | 35 | 8,378 | 1.0M | $0.25 | $1.50 | β | β | β | β | |
| 45 | Qwen: Qwen3.5 Plus 2026-02-15 | qwen | 61.9% | 0.012 | β | $0.0147 | 225 | 9,355 | 1M | $0.26 | $1.56 | β | β | β | β |
| 46 | Mistral: Mistral Large 3 2512 | mistralai | 59.0% | 0.17 | β | $0.0024 | 18 | 1,425 | 262K | $0.50 | $1.50 | β | β | β | β |
| 47 | ByteDance Seed: Seed 1.6 Flash | bytedance-seed | 57.7% | 0.080 | β | $0.0022 | 53 | 7,315 | 262K | $0.07 | $0.30 | β | β | β | β |
| 48 | Mistral: Mistral Small 4 | mistralai | 57.7% | 0.065 | β | $0.0046 | 55 | 6,423 | 262K | $0.15 | $0.60 | Open | β | β | β |
| 49 | Perceptron: Perceptron Mk1 | perceptron | 57.7% | 0.055 | β | $0.0009 | 18 | 569 | 33K | $0.15 | $1.50 | β | β | β | β |
| 50 | Google: Gemini 2.5 Flash Lite | 57.3% | 0.13 | β | $0.0037 | 44 | 9,056 | 1.0M | $0.10 | $0.40 | β | β | β | β | |
| 51 | Qwen: Qwen3 VL 8B Instruct | qwen | 56.9% | 0.028 | β | $0.0022 | 63 | 4,203 | 262K | $0.25 | $0.75 | Open | β | β | β |
| 52 | Google: Gemini 3.1 Flash Lite | 56.5% | 0.13 | 0.021 | $0.0191 | 47 | 12,568 | 1.0M | $0.25 | $1.50 | β | β | 33.5 | 82% | |
| 53 | Mistral: Ministral 3 14B 2512 | mistralai | 56.5% | 0.086 | β | $0.0007 | 25 | 2,286 | 262K | $0.20 | $0.20 | Open | β | β | β |
| 54 | Google: Gemma 4 26B A4B | 55.2% | 0.003 | β | $0.0043 | 258 | 11,369 | 262K | $0.08 | $0.26 | Open | β | β | β | |
| 55 | Tencent: Hy3 preview | tencent | 54.0% | 0.007 | β | $0.0051 | 216 | 20,722 | 262K | $0.18 | $0.60 | Open | β | 41.9 | 87% |
| 56 | Mistral: Ministral 3 8B 2512 | mistralai | 54.0% | 0.065 | β | $0.0005 | 22 | 2,114 | 262K | $0.15 | $0.15 | Open | β | β | β |
| 57 | OpenAI: GPT-5 Nano | openai | 54.0% | 0.026 | β | $0.0047 | 80 | 11,727 | 400K | $0.05 | $0.40 | β | β | 26.8 | 68% |
| 58 | Mistral: Mistral Small 3.2 24B | mistralai | 53.6% | 0.017 | β | $0.0006 | 42 | 2,260 | 256K | $0.07 | $0.20 | Open | β | β | β |
| 59 | Meta: Llama 4 Maverick | meta-llama | 52.7% | 0.008 | β | $0.0010 | 38 | 1,141 | 1.0M | $0.19 | $0.65 | Open | β | 18.4 | 67% |
| 60 | Google: Gemma 3 27B | 50.2% | 0.006 | β | $0.0003 | 32 | 897 | 131K | $0.08 | $0.45 | Open | 1359 | 10.3 | 43% | |
| 61 | Qwen: Qwen2.5 VL 72B Instruct | qwen | 50.2% | 0.005 | β | $0.0009 | 35 | 933 | 128K | $0.80 | $1.00 | Open | β | β | β |
| 62 | StepFun: Step 3.7 Flash | stepfun | 49.8% | 0.005 | β | $0.0162 | 131 | 13,977 | 262K | $0.20 | $1.15 | β | β | 42.6 | 81% |
| 63 | MiniMax: MiniMax-01 | minimax | 49.4% | 0.005 | β | $0.0014 | 27 | 672 | 1.0M | $0.20 | $1.10 | Open | β | β | β |
| 64 | Qwen: Qwen3.5-27B | qwen | 49.4% | <0.001 | β | $0.0348 | 293 | 13,481 | 262K | $0.26 | $2.60 | Open | β | β | β |
| 65 | Google: Gemma 3 12B | 49.0% | 0.009 | β | $0.0002 | 22 | 947 | 131K | $0.05 | $0.15 | Open | 1337 | 8.8 | 35% | |
| 66 | Meta: Llama 4 Scout | meta-llama | 48.5% | 0.009 | β | $0.0005 | 23 | 994 | 1.3M | $0.18 | $0.59 | Open | β | 13.5 | 59% |
| 67 | NVIDIA: Nemotron 3 Nano Omni (free) | nvidia | 48.5% | 0.016 | β | $0.0000 | 54 | 10,809 | 256K | $0.00 | $0.00 | β | β | β | β |
| 68 | OpenAI: GPT-4.1 Nano | openai | 48.1% | 0.044 | β | $0.0008 | 13 | 1,766 | 1.0M | $0.10 | $0.40 | β | 1287 | 13.0 | 51% |
| 69 | ByteDance Seed: Seed-2.0-Lite | bytedance-seed | 47.7% | <0.001 | β | $0.0080 | 285 | 3,821 | 262K | $0.25 | $2.00 | β | β | β | β |
| 70 | OpenAI: GPT-4o-mini | openai | 47.7% | 0.021 | β | $0.0019 | 11 | 682 | 128K | $0.15 | $0.60 | β | β | 12.6 | 43% |
| 71 | Qwen: Qwen3.6 35B A3B | qwen | 47.7% | 0.004 | β | $0.0189 | 124 | 15,382 | 262K | $0.15 | $1.00 | Open | β | 43.5 | 84% |
| 72 | Google: Gemma 3 4B | 47.3% | 0.003 | β | $0.0001 | 36 | 1,081 | 131K | $0.05 | $0.10 | Open | 1291 | 6.3 | 29% | |
| 73 | Mistral: Ministral 3 3B 2512 | mistralai | 45.6% | 0.021 | β | $0.0004 | 18 | 2,863 | 131K | $0.10 | $0.10 | Open | β | β | β |
| 74 | Perplexity: Sonar | perplexity | 45.6% | 0.004 | β | $0.0119 | 6.93 | 75 | 127K | $1.00 | $1.00 | β | β | β | β |
| 75 | Qwen: Qwen3.6 Plus | qwen | 44.8% | <0.001 | β | $0.0138 | 321 | 7,030 | 1M | $0.33 | $1.95 | β | β | 50.0 | 88% |
| 76 | NVIDIA: Nemotron Nano 12B 2 VL (free) | nvidia | 42.3% | 0.002 | β | $0.0000 | 15 | 465 | 128K | $0.00 | $0.00 | Open | β | β | β |
| 77 | Anthropic: Claude 3 Haiku | anthropic | 41.8% | 0.014 | β | $0.0009 | 5.77 | 519 | 200K | $0.25 | $1.25 | β | 1194 | β | β |
| 78 | Arcee AI: Trinity Large Thinking | arcee-ai | 41.6% | <0.001 | β | $0.0108 | 102 | 11,608 | 262K | $0.25 | $0.80 | Open | β | 31.9 | 75% |
| 79 | Qwen: Qwen3.5-9B | qwen | 41.4% | <0.001 | β | $0.0026 | 476 | 14,442 | 262K | $0.10 | $0.15 | Open | β | 32.4 | 81% |
| 80 | Amazon: Nova Lite 1.0 | amazon | 39.7% | 0.012 | β | $0.0002 | 5.31 | 666 | 300K | $0.06 | $0.24 | β | β | β | β |
| 81 | Qwen: Qwen3.5-35B-A3B | qwen | 39.3% | <0.001 | β | $0.0228 | 290 | 22,255 | 262K | $0.15 | $1.00 | Open | β | β | β |
| 82 | ByteDance Seed: Seed-2.0-Mini | bytedance-seed | 37.2% | <0.001 | β | $0.0013 | 382 | 2,871 | 262K | $0.10 | $0.40 | β | β | β | β |
| 83 | Meta: Llama 3.2 11B Vision Instruct | meta-llama | 35.6% | <0.001 | β | $0.0023 | 74 | 3,340 | 131K | $0.34 | $0.34 | Open | β | β | β |
Showing 82 of 551 models Β· as of 2026-10-18. Methodology & π β
Third-party benchmark values: Artificial Analysis (artificialanalysis.ai) β shown under their attribution terms; not redistributed in downloads.
Crowd preference (Arena Elo): LMArena / Chatbot Arena via mathewhe/chatbot-arena-elo (Apache-2.0).
Model metadata + pricing via OpenRouter. First-party scores + Effective Value (π) by the meo-benchmark project.
Embed this leaderboard
Drop the live top-10 on your site β it always shows current data and links back here.