Comprehensive ranking of the top 50 cutting-edge AI models based on Neural Processing and Verbal Articulation (NPOVA) benchmarks and writing proficiency.
| Rank | Model | Organization | License | NPOVA Score | Write Score (%) | Details |
|---|---|---|---|---|---|---|
| 1 🥇 | Gemini-2.5-Pro-Preview-05-06 |
Proprietary | 96.82 | 90.23 | ||
| 2 🥈 | Claude 3.7 Sonnet (20250219) |
Anthropic | Proprietary | 95.10 | 89.90 | |
| 3 🥉 | Gemini-2.5-Flash-Preview-05-20 |
Proprietary | 93.01 | 89.30 | ||
| 4 | GPT-4.1-2025-04-14 | OpenAI | Proprietary | 91.28 | 88.50 | |
| 5 | Claude 3.5 Sonnet (20241022) | Anthropic | Proprietary | 90.00 | 87.70 | |
| 6 | DeepSeek-V3-0324 | DeepSeek AI | MIT | 88.70 | 86.90 | |
| 7 | DeepSeek-R1 | DeepSeek AI | MIT | 87.80 | 86.10 | |
| 8 | o3-2025-04-16 | OpenAI | Proprietary | 87.20 | 85.70 | |
| 9 | GPT-4.1-mini-2025-04-14 | OpenAI | Proprietary | 86.40 | 85.00 | |
| 10 | Qwen3-235B-A22B | Alibaba Cloud | Apache 2.0 | 85.50 | 84.20 | |
| 11 | Mistral Medium 3 | Mistral AI | Proprietary | 84.10 | 83.30 | |
| 12 | early-grok-3 | xAI | Proprietary | 83.20 | 82.80 | |
| 13 | Gemini-2.5-Flash-Preview-04-17 | Proprietary | 83.10 | 82.60 | ||
| 14 | o3-mini-high (20250131) | OpenAI | Proprietary | 82.80 | 82.00 | |
| 15 | Claude 3.5 Haiku (20241022) | Anthropic | Proprietary | 82.40 | 81.50 | |
| 16 | o4-mini-2025-04-16 | OpenAI | Proprietary | 81.10 | 80.10 | |
| 17 | o3-mini (20250131) | OpenAI | Proprietary | 80.80 | 79.80 | |
| 18 | Gemini-2.0-Pro-Exp-02-05 | Proprietary | 80.40 | 79.50 | ||
| 19 | o1 (20241217) | OpenAI | Proprietary | 78.30 | 78.00 | |
| 20 | o1-mini (20240912) | OpenAI | Proprietary | 78.00 | 77.60 | |
| 21 | Gemini-2.0-Flash-001 | Proprietary | 77.80 | 77.20 | ||
| 22 | Gemini-2.0-Flash-Thinking-01-21 | Proprietary | 76.80 | 76.00 | ||
| 23 | Llama-4-Maverick-17B | Meta | Llama 4 | 75.90 | 75.00 | |
| 24 | Zephyr Omega-70B | HuggingFace | Apache 2.0 | 75.20 | 74.50 | |
| 25 | Command R+ (Gen 2) | Cohere | Proprietary | 74.80 | 74.10 | |
| 26 | Yi-Large-Turbo | 01.AI | Proprietary | 74.50 | 73.80 | |
| 27 | Orion-14B-ChatMax | OrionLM | Orion License | 74.10 | 73.50 | |
| 28 | DBRX Instruct v2 | Databricks | Databricks Open | 73.80 | 73.00 | |
| 29 | Gemma-2-27B-Evo | Gemma T&C | 73.50 | 72.80 | ||
| 30 | Mistral Large 2 (Exp) | Mistral AI | Proprietary | 73.00 | 72.50 | |
| 31 | Phi-3-Vision-128k | Microsoft | MIT | 72.60 | 72.00 | |
| 32 | Jamba-Instruct-Pro | AI21 Labs | Proprietary | 72.20 | 71.50 | |
| 33 | Starling-LM-7B-beta | Nexusflow | Apache 2.0 | 71.80 | 71.00 | |
| 34 | WizardLM-2-8x22B | Microsoft | WizardLM | 71.50 | 70.80 | |
| 35 | Arctic-Instruct | Snowflake | Apache 2.0 | 71.00 | 70.20 | |
| 36 | DeepSeek-Coder-V2 Lite | DeepSeek AI | MIT | 70.60 | 69.80 | |
| 37 | Qwen2-72B-Instruct | Alibaba Cloud | Tongyi Qianwen | 70.20 | 69.50 | |
| 38 | Llama-3-70B-Instruct-Plus | Meta | Llama 3 | 69.80 | 69.00 | |
| 39 | Falcon-180B-Chat (Adv) | TII | Apache 2.0 Mod | 69.40 | 68.60 | |
| 40 | Nemotron-4-340B-Base | NVIDIA | Nemotron | 69.00 | 68.20 | |
| 41 | OLMo-7B-Instruct-Pro | AI2 | Apache 2.0 | 68.50 | 67.80 | |
| 42 | XVerse-70B-Chat-Plus | Shenzhen Transsion | XVERSE | 68.10 | 67.30 | |
| 43 | SeaLLM-7B-v2.5-Instruct | Sea AI Labs | MIT | 67.70 | 66.90 | |
| 44 | InternLM2-Chat-20B-Pro | Shanghai AI Lab | Apache 2.0 | 67.20 | 66.50 | |
| 45 | MPT-30B-Chat-Plus | MosaicML | Apache 2.0 | 66.80 | 66.00 | |
| 46 | OpenChat-3.5-1210-Pro | OpenChat | Apache 2.0 | 66.30 | 65.50 | |
| 47 | Vicuna-33B-v1.3-Adv | LMSYS | Apache 2.0 | 65.90 | 65.10 | |
| 48 | CodeLlama-70B-Instruct-HF | Meta | Llama Community | 65.40 | 64.70 | |
| 49 | StableLM-Zephyr-3B-Plus | Stability AI | CreativeML Open | 65.00 | 64.20 | |
| 50 | SOLAR-10.7B-Instruct-v1.0 | Upstage | Apache 2.0 | 64.60 | 63.80 |