diff --git a/README.md b/README.md index d7a750a6..5c254f8a 100644 --- a/README.md +++ b/README.md @@ -32,23 +32,23 @@ RouterArena bridges this gap by providing an open evaluation platform and benchm For more details, please see our [website](https://routeworks.github.io/leaderboard) and [blog](https://huggingface.co/blog/JerryPotter/who-routes-the-routers). -| Rank | Router | Affiliation | Arena | Opt.Sel | Opt.Cost | Opt.Acc | Lat | Robust | -|------|---------|--------------|--------|------------|-----------|----------|----------|-------------| -| ๐Ÿฅ‡ | MIRT-BERT | ๐ŸŽ“ USTC | 66.89 | 3.44 | 19.62 | 78.18 | 27.03 | 94.50 | -| ๐Ÿฅˆ | Azure | ๐Ÿ’ผ Microsoft | 66.66 | 22.52 | 46.32 | 81.96 | โ€” | โ€” | -| ๐Ÿฅ‰ | NIRT-BERT | ๐ŸŽ“ USTC | 66.12 | 3.83 | 14.04 | 77.88 | 10.42 | 44.50 | -| 4 | GPT-5 | ๐Ÿ’ผ OpenAI | 64.32 | โ€” | โ€” | โ€” | โ€” | โ€” | -| 5 | vLLM-SR | ๐Ÿ’ผ vLLM | 64.32 | 4.79 | 12.54 | 79.33 | 0.19 | 100.00 | -| 6 | CARROT | ๐ŸŽ“ UMich | 63.87 | 2.68 | 6.77 | 78.63 | 1.50 | 93.60 | -| 7 | Chayan | ๐Ÿ’ผ Adaptive Classifier | 63.83 | 43.03 | 43.75 | 88.74 | โ€” | โ€” | -| 8 | NotDiamond | ๐Ÿ’ผ NotDiamond | 63.00 | 1.55 | 2.14 | 76.81 | โ€” | โ€” | -| 9 | MLP | ๐ŸŽ“ Academic | 57.56 | 13.39 | 24.45 | 83.32 | 90.91 | 96.90 | -| 10 | GraphRouter | ๐ŸŽ“ UIUC | 57.22 | 4.73 | 38.33 | 74.25 | 2.70 | 97.50 | -| 11 | KNN | ๐ŸŽ“ Academic | 55.48 | 13.09 | 25.49 | 78.77 | 1.33 | 51.30 | -| 12 | RouteLLM | ๐ŸŽ“ Berkeley | 48.07 | 99.72 | 99.63 | 68.76 | 0.40 | 99.80 | -| 13 | RouterDC | ๐ŸŽ“ SUSTech | 33.75 | 39.84 | 73.00 | 49.05 | 10.75 | 97.60 | - -๐ŸŽ“ Academicโ€ƒโ€ƒ๐Ÿ’ผ Commercialโ€ƒ +| Rank | Router | Affiliation | Arena | Optimal Selection | Optimal Cost | Optimal Accuracy | Latency | Robustness | +|------|--------------------|-----------------------------|--------|-----------------|--------------|----------------|---------|------------| +| ๐Ÿฅ‡ | [MIRTโ€‘BERT](https://arxiv.org/pdf/2506.01048) [[GH]](https://github.com/Mercidaiha/IRT-Router) | ๐ŸŽ“ USTC | 66.89 | 3.44 | 19.62 | 78.18 | 27.03 | 94.50 | +| ๐Ÿฅˆ | [Azureโ€‘Router](https://ai.azure.com/catalog/models/model-router) [[Web]](https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/model-router) | ๐Ÿ’ผ Microsoft | 66.66 | 22.52 | 46.32 | 81.96 | โ€” | โ€” | +| ๐Ÿฅ‰ | [NIRTโ€‘BERT](https://arxiv.org/pdf/2506.01048) [[GH]](https://github.com/Mercidaiha/IRT-Router) | ๐ŸŽ“ USTC | 66.12 | 3.83 | 14.04 | 77.88 | 10.42 | 44.50 | +| 4 | [GPTโ€‘5](https://openai.com/index/introducing-gpt-5/)| ๐Ÿ’ผ OpenAI | 64.32 | โ€” | โ€” | โ€” | โ€” | โ€” | +| 5 | [vLLMโ€‘SR](https://vllm-semantic-router.com/) [[GH]](https://github.com/vllm-project/semantic-router) [[HF]](https://huggingface.co/llm-semantic-router) | ๐Ÿ’ผ vLLM | 64.32 | 4.79 | 12.54 | 79.33 | 0.19 | 100.00 | +| 6 | [CARROT](https://arxiv.org/abs/2502.03261) [[GH]](https://github.com/somerstep/CARROT) [[HF]](https://huggingface.co/CARROT-LLM-Routing) | ๐ŸŽ“ UMich | 63.87 | 2.68 | 6.77 | 78.63 | 1.50 | 93.60 | +| 7 | [Chayan](https://huggingface.co/adaptive-classifier/chayan) [[HF]](https://huggingface.co/adaptive-classifier/chayan) | ๐ŸŽ“ Adaptive Classifier | 63.83 | 43.03 | 43.75 | 88.74 | โ€” | โ€” | +| 8 | [NotDiamond](https://www.notdiamond.ai/) | ๐Ÿ’ผ NotDiamond | 63.00 | 1.55 | 2.14 | 76.81 | โ€” | โ€” | +| 9 | [RouterBenchโ€‘MLP](https://arxiv.org/pdf/2403.12031) [[GH]](https://github.com/withmartian/routerbench) [[HF]](https://huggingface.co/datasets/withmartian/routerbench) | ๐ŸŽ“ Martian | 57.56 | 13.39 | 24.45 | 83.32 | 90.91 | 96.90 | +| 10 | [GraphRouter](https://arxiv.org/abs/2410.03834) [[GH]](https://github.com/ulab-uiuc/GraphRouter) | ๐ŸŽ“ UIUC | 57.22 | 4.73 | 38.33 | 74.25 | 2.70 | 97.50 | +| 11 | [RouterBenchโ€‘KNN](https://arxiv.org/pdf/2403.12031) [[GH]](https://github.com/withmartian/routerbench) [[HF]](https://huggingface.co/datasets/withmartian/routerbench) | ๐ŸŽ“ Martian | 55.48 | 13.09 | 25.49 | 78.77 | 1.33 | 51.30 | +| 12 | [RouteLLM](https://arxiv.org/abs/2406.18665) [[GH]](https://github.com/lm-sys/RouteLLM) [[HF]](https://huggingface.co/routellm) | ๐ŸŽ“ Berkeley | 48.07 | 99.72 | 99.63 | 68.76 | 0.40 | 99.80 | +| 13 | [RouterDC](https://arxiv.org/abs/2409.19886) [[GH]](https://github.com/shuhao02/RouterDC) | ๐ŸŽ“ SUSTech | 33.75 | 39.84 | 73.00 | 49.05 | 10.75 | 97.60 | + +๐ŸŽ“ Open-sourceโ€ƒโ€ƒ๐Ÿ’ผ Closed-sourceโ€ƒ