Skip to content

Yl231 sync leaderboard 2026 05 #40

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
yl231 merged 4 commits into from
May 20, 2026
Merged

Yl231 sync leaderboard 2026 05 #40

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
1d2c6c2
Sync leaderboard with RouterArena README (Weave, Auto, R2 entries)
yl231 May 20, 2026
fb334ff
Add Sqwish Router entry; default null per-domain cells to 0
yl231 May 20, 2026
c16a1cb
Apply PR #97 optimality fix to all workflow-evaluated routers
yl231 May 20, 2026
6aee9b4
Format data files with prettier (trailing newlines, inline single-ele…
yl231 May 20, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1,682 changes: 1,682 additions & 0 deletions src/data/flip_labels/flip_labels_auto_router.json
View file
Open in desktop

Large diffs are not rendered by default.

1,682 changes: 1,682 additions & 0 deletions src/data/flip_labels/flip_labels_r2-router.json
View file
Open in desktop

Large diffs are not rendered by default.

1,682 changes: 1,682 additions & 0 deletions src/data/flip_labels/flip_labels_sqwish-router.json
View file
Open in desktop

Large diffs are not rendered by default.

1,682 changes: 1,682 additions & 0 deletions src/data/flip_labels/flip_labels_vllm-sr.json
View file
Open in desktop

Large diffs are not rendered by default.

1,682 changes: 1,682 additions & 0 deletions src/data/flip_labels/flip_labels_weave-router.json
View file
Open in desktop

Large diffs are not rendered by default.

3,728 changes: 3,728 additions & 0 deletions src/data/routerMetrics/category_scores.json
View file
Open in desktop

Large diffs are not rendered by default.

204 changes: 113 additions & 91 deletions src/data/routerMetrics/leaderboard.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,36 +1,69 @@
[
{
"Router Name": "Sqwish Router",
"Arena Score": 75.27,
"Optimal Selection Score": 7.41,
"Optimal Cost Score": 25.1,
"Optimal Acc. Score": 90.47,
"Robustness Score": 100.0,
"Latency Score": null,
"Accuracy": 76.4,
"Cost per 1k": 0.18
},
{
"Router Name": "Weave Router",
"Arena Score": 74.61,
"Optimal Selection Score": 1.38,
"Optimal Cost Score": 12.27,
"Optimal Acc. Score": 100.00,
"Optimal Selection Score": 1.27,
"Optimal Cost Score": 12.1,
"Optimal Acc. Score": 89.56,
"Robustness Score": 79.05,
"Latency Score": null,
"Accuracy": 78.43,
"Cost per 1k": 0.92
},
{
"Router Name": "R2-Router",
"Arena Score": 71.60,
"Optimal Selection Score": 32.38,
"Optimal Cost Score": 74.16,
"Optimal Acc. Score": 100.00,
"Arena Score": 71.6,
"Optimal Selection Score": 24.51,
"Optimal Cost Score": 48.7,
"Optimal Acc. Score": 99.85,
"Robustness Score": 45.71,
"Latency Score": null,
"Accuracy": 71.23,
"Cost per 1k": 0.06
},
{
"Router Name": "RouterDC",
"Arena Score": 33.75,
"Optimal Selection Score": 39.84,
"Optimal Cost Score": 72.998,
"Optimal Acc. Score": 49.05,
"Robustness Score": 85.24,
"Latency Score": 10.7527,
"Accuracy": 32.01,
"Cost per 1k": 0.07
"Router Name": "Auto Router",
"Arena Score": 70.05,
"Optimal Selection Score": 37.58,
"Optimal Cost Score": 40.02,
"Optimal Acc. Score": 86.04,
"Robustness Score": 49.52,
"Latency Score": null,
"Accuracy": 70.17,
"Cost per 1k": 0.12
},
{
"Router Name": "vllm",
"Arena Score": 67.23,
"Optimal Selection Score": 84.66,
"Optimal Cost Score": 90.71,
"Optimal Acc. Score": 89.24,
"Robustness Score": 90.95,
"Latency Score": null,
"Accuracy": 66.53,
"Cost per 1k": 0.06
},
{
"Router Name": "mirt_bert",
"Arena Score": 66.89,
"Optimal Selection Score": 3.44,
"Optimal Cost Score": 19.6178,
"Optimal Acc. Score": 78.18,
"Robustness Score": 61.19,
"Latency Score": 27.027,
"Accuracy": 66.88,
"Cost per 1k": 0.15
},
{
"Router Name": "azure",
Expand All @@ -44,15 +77,15 @@
"Cost per 1k": 0.54
},
{
"Router Name": "carrot",
"Arena Score": 63.87,
"Optimal Selection Score": 2.68,
"Optimal Cost Score": 6.7697,
"Optimal Acc. Score": 78.63,
"Robustness Score": 89.05,
"Latency Score": 1.4993,
"Accuracy": 67.21,
"Cost per 1k": 2.06
"Router Name": "nirt_bert",
"Arena Score": 66.12,
"Optimal Selection Score": 3.83,
"Optimal Cost Score": 14.039,
"Optimal Acc. Score": 77.88,
"Robustness Score": 49.29,
"Latency Score": 10.4167,
"Accuracy": 66.34,
"Cost per 1k": 0.21
},
{
"Router Name": "gpt5",
Expand All @@ -66,41 +99,41 @@
"Cost per 1k": 10.02
},
{
"Router Name": "graphrouter",
"Arena Score": 57.22,
"Optimal Selection Score": 4.73,
"Optimal Cost Score": 38.3347,
"Optimal Acc. Score": 74.25,
"Robustness Score": 94.29,
"Latency Score": 2.6954,
"Accuracy": 57.0,
"Cost per 1k": 0.34
"Router Name": "carrot",
"Arena Score": 63.87,
"Optimal Selection Score": 2.68,
"Optimal Cost Score": 6.7697,
"Optimal Acc. Score": 78.63,
"Robustness Score": 89.05,
"Latency Score": 1.4993,
"Accuracy": 67.21,
"Cost per 1k": 2.06
},
{
"Router Name": "mirt_bert",
"Arena Score": 66.89,
"Optimal Selection Score": 3.44,
"Optimal Cost Score": 19.6178,
"Optimal Acc. Score": 78.18,
"Robustness Score": 61.19,
"Latency Score": 27.027,
"Accuracy": 66.88,
"Cost per 1k": 0.15
"Router Name": "chayan",
"Arena Score": 63.83,
"Optimal Selection Score": 43.03,
"Optimal Cost Score": 43.75,
"Optimal Acc. Score": 88.74,
"Robustness Score": null,
"Latency Score": null,
"Accuracy": 64.89,
"Cost per 1k": 0.56
},
{
"Router Name": "nirt_bert",
"Arena Score": 66.12,
"Optimal Selection Score": 3.83,
"Optimal Cost Score": 14.039,
"Optimal Acc. Score": 77.88,
"Robustness Score": 49.29,
"Latency Score": 10.4167,
"Accuracy": 66.34,
"Cost per 1k": 0.21
"Router Name": "routerbench_mlp",
"Arena Score": 57.56,
"Optimal Selection Score": 13.39,
"Optimal Cost Score": 24.4499,
"Optimal Acc. Score": 83.32,
"Robustness Score": 80.0,
"Latency Score": 90.9091,
"Accuracy": 61.62,
"Cost per 1k": 4.83
},
{
"Router Name": "notdiamond",
"Arena Score": 63.0,
"Arena Score": 57.29,
"Optimal Selection Score": 1.55,
"Optimal Cost Score": 2.1367,
"Optimal Acc. Score": 76.81,
Expand All @@ -110,15 +143,15 @@
"Cost per 1k": 4.1
},
{
"Router Name": "routellm",
"Arena Score": 48.07,
"Optimal Selection Score": 99.72,
"Optimal Cost Score": 99.6314,
"Optimal Acc. Score": 68.76,
"Robustness Score": 100.00,
"Latency Score": 0.4016,
"Accuracy": 47.04,
"Cost per 1k": 0.27
"Router Name": "graphrouter",
"Arena Score": 57.22,
"Optimal Selection Score": 4.73,
"Optimal Cost Score": 38.3347,
"Optimal Acc. Score": 74.25,
"Robustness Score": 94.29,
"Latency Score": 2.6954,
"Accuracy": 57.0,
"Cost per 1k": 0.34
},
{
"Router Name": "routerbench_knn",
Expand All @@ -132,36 +165,25 @@
"Cost per 1k": 4.27
},
{
"Router Name": "routerbench_mlp",
"Arena Score": 57.56,
"Optimal Selection Score": 13.39,
"Optimal Cost Score": 24.4499,
"Optimal Acc. Score": 83.32,
"Robustness Score": 80.00,
"Latency Score": 90.9091,
"Accuracy": 61.62,
"Cost per 1k": 4.83
},
{
"Router Name": "vllm",
"Arena Score": 67.23,
"Optimal Selection Score": 94.10,
"Optimal Cost Score": 90.12,
"Optimal Acc. Score": 100.00,
"Robustness Score": 90.95,
"Latency Score": null,
"Accuracy": 66.53,
"Cost per 1k": 0.06
"Router Name": "routellm",
"Arena Score": 48.07,
"Optimal Selection Score": 99.72,
"Optimal Cost Score": 99.6314,
"Optimal Acc. Score": 68.76,
"Robustness Score": 100.0,
"Latency Score": 0.4016,
"Accuracy": 47.04,
"Cost per 1k": 0.27
},
{
"Router Name": "chayan",
"Arena Score": 63.83,
"Optimal Selection Score": 43.03,
"Optimal Cost Score": 43.75,
"Optimal Acc. Score": 88.74,
"Robustness Score": null,
"Latency Score": null,
"Accuracy": 64.89,
"Cost per 1k": 0.56
"Router Name": "RouterDC",
"Arena Score": 33.75,
"Optimal Selection Score": 39.84,
"Optimal Cost Score": 72.998,
"Optimal Acc. Score": 49.05,
"Robustness Score": 85.24,
"Latency Score": 10.7527,
"Accuracy": 32.01,
"Cost per 1k": 0.07
}
]
55 changes: 26 additions & 29 deletions src/data/routers.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,17 @@
{
"Sqwish Router": {
"name": "Sqwish Router",
"type": "closed-source",
"description": "Sqwish Router final submission.",
"affiliation": "Sqwish",
"modelPool": [
"qwen/qwen3-235b-a22b-2507",
"qwen/qwen3-next-80b-a3b-instruct",
"Qwen/Qwen3-Coder-Next",
"google/gemini-3.1-flash-lite",
"deepseek/deepseek-v4-flash"
]
},
"Weave Router": {
"name": "Weave Router",
"type": "open-source",
Expand Down Expand Up @@ -128,27 +141,6 @@
"websiteUrl": "https://arxiv.org/abs/2410.03834",
"githubUrl": "https://github.com/ulab-uiuc/GraphRouter"
},
"universal_router": {
"name": "Universal Router",
"type": "open-source",
"description": "Universal baseline that mirrors the RouterBench model mix",
"affiliation": "RouterArena",
"modelPool": [
"WizardLM/WizardLM-13B-V1.2",
"claude-instant-v1",
"claude-v1",
"claude-v2",
"gpt-3.5-turbo-1106",
"gpt-4-1106-preview",
"meta/codellama-34b-instruct",
"meta/llama-2-70b-chat",
"mistralai/mistral-7b-chat",
"mistralai/mixtral-8x7b-chat",
"zero-one-ai/Yi-34B-Chat"
],
"paperUrl": "https://huggingface.co/datasets/withmartian/routerbench",
"websiteUrl": "https://huggingface.co/datasets/withmartian/routerbench"
},
"mirt_bert": {
"name": "MIRT-BERT",
"type": "open-source",
Expand Down Expand Up @@ -226,26 +218,22 @@
"openai/gpt-4.1-mini",
"openai/gpt-4.1-nano",
"openai/o1-mini",

"anthropic/claude-3-7-sonnet-20250219",
"anthropic/claude-3-5-haiku-20241022",
"anthropic/claude-3-haiku-20240307",
"anthropic/claude-sonnet-4-20250514",

"google/gemini-1.5-pro-latest",
"google/gemini-1.5-flash-latest",
"google/gemini-2.0-flash-001",
"google/gemini-2.5-flash",
"google/gemini-2.5-pro",

"mistral/codestral-latest",
"mistral/open-mixtral-8x7b",
"mistral/mistral-large-latest",
"mistral/mistral-medium-latest",
"mistral/mistral-small-latest",
"mistral/open-mistral-7b",
"mistral/open-mistral-nemo",

"togetherai/Meta-Llama-3.1-8B-Instruct-Turbo",
"togetherai/Meta-Llama-3.1-70B-Instruct-Turbo",
"togetherai/Meta-Llama-3.1-405B-Instruct-Turbo"
Expand All @@ -258,10 +246,7 @@
"type": "open-source",
"description": "Binary selection between strong and weak models",
"affiliation": "Berkeley",
"modelPool": [
"openai-gpt-4o",
"mixtral-8x7b-instruct"
],
"modelPool": ["openai-gpt-4o", "mixtral-8x7b-instruct"],
"paperUrl": "https://arxiv.org/abs/2406.18665",
"websiteUrl": "https://arxiv.org/abs/2406.18665",
"githubUrl": "https://github.com/lm-sys/RouteLLM",
Expand Down Expand Up @@ -347,5 +332,17 @@
"paperUrl": "https://huggingface.co/adaptive-classifier/chayan",
"websiteUrl": "https://huggingface.co/adaptive-classifier/chayan",
"huggingfaceUrl": "https://huggingface.co/adaptive-classifier/chayan"
},
"Auto Router": {
"name": "Auto Router",
"type": "open-source",
"description": "Cyclic baseline that rotates through its 4-model pool.",
"affiliation": "RouterArena",
"modelPool": [
"deepseek/deepseek-v3.2",
"qwen/qwen3-235b-a22b-2507",
"qwen/qwen3.5-9b",
"openai/gpt-4o"
]
}
}
Loading
Loading