| Model,MT-bench (score),Arena Elo rating,MMLU,License,Link | |
| GPT-4,8.99,1211,0.864,Proprietary,https://openai.com/research/gpt-4 | |
| Claude-2,8.06,-,0.785,Proprietary,https://www.anthropic.com/index/claude-2 | |
| GPT-3.5-turbo,7.94,1124,0.700,Proprietary,https://openai.com/blog/chatgpt | |
| Claude-v1,7.90,1169,0.770,Proprietary,https://www.anthropic.com/index/introducing-claude | |
| Claude-instant-v1,7.85,1145,0.734,Proprietary,https://www.anthropic.com/index/introducing-claude | |
| Vicuna-33B,7.12,1096,0.592,Non-commercial,https://huggingface.co./lmsys/vicuna-33b-v1.3 | |
| WizardLM-30B,7.01,-,0.587,Non-commercial,https://huggingface.co./WizardLM/WizardLM-30B-V1.0 | |
| Llama-2-70b-chat,6.86,-,-,Llama 2 Community,https://huggingface.co./meta-llama/Llama-2-70b-hf | |
| WizardLM-13B-v1.1,6.76,-,-,Non-commercial,https://huggingface.co./WizardLM/WizardLM-13B-V1.1 | |
| Llama-2-13b-chat,6.65,-,-,Llama 2 Community,https://huggingface.co./meta-llama/Llama-2-13b-hf | |
| Guanaco-33B,6.53,1044,0.576,Non-commercial,https://huggingface.co./timdettmers/guanaco-33b-merged | |
| Tulu-30B,6.43,-,0.581,Non-commercial,https://huggingface.co./allenai/tulu-30b | |
| Guanaco-65B,6.41,-,0.621,Non-commercial,https://huggingface.co./timdettmers/guanaco-65b-merged | |
| OpenAssistant-LLaMA-30B,6.41,-,0.560,Non-commercial,https://huggingface.co./OpenAssistant/oasst-sft-6-llama-30b-xor | |
| PaLM-Chat-Bison-001,6.40,1019,-,Proprietary,https://cloud.google.com/vertex-ai/docs/generative-ai/learn/models#foundation_models | |
| Vicuna-13B,6.39,1055,0.521,Non-commercial,https://huggingface.co./lmsys/vicuna-13b-v1.3 | |
| MPT-30B-chat,6.39,1049,0.504,CC-BY-NC-SA-4.0,https://huggingface.co./mosaicml/mpt-30b-chat | |
| WizardLM-13B,6.35,1043,0.523,Non-commercial,https://huggingface.co./WizardLM/WizardLM-13B-V1.0 | |
| Llama-2-7b-chat,6.27,-,-,Llama 2 Community,https://huggingface.co./meta-llama/Llama-2-7b-chat-hf | |
| Vicuna-7B,6.00,1006,0.471,Non-commercial,https://huggingface.co./lmsys/vicuna-7b-v1.3 | |
| Baize-v2-13B,5.75,-,0.489,Non-commercial,https://huggingface.co./project-baize/baize-v2-13b | |
| XGen-7B-8K-Inst,5.55,-,-,Non-commercial,https://huggingface.co./Salesforce/xgen-7b-8k-inst | |
| Nous-Hermes-13B,5.51,-,0.493,Non-commercial,https://huggingface.co./NousResearch/Nous-Hermes-13b | |
| MPT-7B-Chat,5.42,951,0.320,CC-BY-NC-SA-4.0,https://huggingface.co./mosaicml/mpt-7b-chat | |
| GPT4All-13B-Snoozy,5.41,971,0.430,Non-commercial,https://huggingface.co./nomic-ai/gpt4all-13b-snoozy | |
| Koala-13B,5.35,987,0.447,Non-commercial,https://bair.berkeley.edu/blog/2023/04/03/koala/ | |
| MPT-30B-Instruct,5.22,-,0.478,CC-BY-SA 3.0,https://huggingface.co./mosaicml/mpt-30b-instruct | |
| Falcon-40B-Instruct,5.17,-,0.547,Apache 2.0,https://huggingface.co./tiiuae/falcon-40b-instruct | |
| ChatGLM2-6B,4.96,-,-,Apache-2.0,https://huggingface.co./THUDM/chatglm2-6b | |
| H2O-Oasst-OpenLLaMA-13B,4.63,-,0.428,Apache 2.0,https://huggingface.co./h2oai/h2ogpt-gm-oasst1-en-2048-open-llama-13b | |
| Alpaca-13B,4.53,926,0.481,Non-commercial,https://crfm.stanford.edu/2023/03/13/alpaca.html | |
| ChatGLM-6B,4.50,904,0.361,Non-commercial,https://huggingface.co./THUDM/chatglm-6b | |
| OpenAssistant-Pythia-12B,4.32,919,0.270,Apache 2.0,https://huggingface.co./OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5 | |
| RWKV-4-Raven-14B,3.98,946,0.256,Apache 2.0,https://huggingface.co./BlinkDL/rwkv-4-raven | |
| Dolly-V2-12B,3.28,846,0.257,MIT,https://huggingface.co./databricks/dolly-v2-12b | |
| FastChat-T5-3B,3.04,897,0.477,Apache 2.0,https://huggingface.co./lmsys/fastchat-t5-3b-v1.0 | |
| StableLM-Tuned-Alpha-7B,2.75,867,0.244,CC-BY-NC-SA-4.0,https://huggingface.co./stabilityai/stablelm-tuned-alpha-7b | |
| LLaMA-13B,2.61,821,0.470,Non-commercial,https://arxiv.org/abs/2302.13971 |