Spaces:

open-agent-leaderboard
/

leaderboard

Running

App Files Files Community

Elron commited on 15 days ago

Commit

a7dde89

verified ·

1 Parent(s): 22f581d

Auto-deploy from GitHub

Browse files

Files changed (7) hide show

README.md +1 -3
index.html +1 -2
results.csv +148 -88
results.csv.timestamp +1 -1
scripts/leaderboard.js +27 -7
scripts/pareto.js +0 -386
styles.css +3 -1

README.md CHANGED Viewed

@@ -14,7 +14,5 @@ Interactive leaderboard and efficiency analysis for general-purpose AI agents ev
 - **Benchmarks**: AppWorld, BrowseComp+, SWE-bench, TauBench (Airline, Retail, Telecom)
 - **Paper**: [arXiv:2602.22953](https://arxiv.org/abs/2602.22953)
-- **Framework**: [Exgentic](https://github.com/Exgentic/exgentic)
-- **GitHub**: [Exgentic/open-agent-leaderboard](https://github.com/Exgentic/open-agent-leaderboard)
 - **Website**: [exgentic.github.io](https://exgentic.github.io)
-- **Submit Results**: [CONTRIBUTING.md](https://github.com/Exgentic/open-agent-leaderboard/blob/main/CONTRIBUTING.md)

 - **Benchmarks**: AppWorld, BrowseComp+, SWE-bench, TauBench (Airline, Retail, Telecom)
 - **Paper**: [arXiv:2602.22953](https://arxiv.org/abs/2602.22953)
+- **GitHub**: [Exgentic/exgentic](https://github.com/Exgentic/exgentic)
 - **Website**: [exgentic.github.io](https://exgentic.github.io)

index.html CHANGED Viewed

@@ -1,6 +1,6 @@
 <!--
 SPDX-License-Identifier: Apache-2.0
-Copyright (C) 2025-2026, The Exgentic organization and its contributors.
 -->
 <!doctype html>
 <html lang="en">
@@ -82,7 +82,6 @@ const header=document.getElementById('header');
 window.addEventListener('scroll',()=>{header.classList.toggle('scrolled',window.scrollY>50)},{passive:true});
 document.getElementById('mobileToggle').addEventListener('click',()=>{document.getElementById('headerNav').classList.toggle('open')});
 </script>
-<script src="./scripts/pareto.js"></script>
 <script src="./scripts/leaderboard.js" data-base="./"></script>
 <script>
 // Theme toggle

 <!--
 SPDX-License-Identifier: Apache-2.0
+Copyright (C) 2025, The Exgentic organization and its contributors.
 -->
 <!doctype html>
 <html lang="en">
 window.addEventListener('scroll',()=>{header.classList.toggle('scrolled',window.scrollY>50)},{passive:true});
 document.getElementById('mobileToggle').addEventListener('click',()=>{document.getElementById('headerNav').classList.toggle('open')});
 </script>
 <script src="./scripts/leaderboard.js" data-base="./"></script>
 <script>
 // Theme toggle

results.csv CHANGED Viewed

@@ -1,91 +1,151 @@
 agent,agent_normalized,visible_agent_name,agent_version,avg_cost,avg_steps,benchmark,finished_pct,model,model_normalized,num_tasks,score,total_cost
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,13.08379215,49.69,AppWorld,0.74,claude,claude-opus-4.5,100,0.66,1308.379215
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0,0,AppWorld,0,gpt52,gpt-5.2,100,0,0
 claudecode,claude-code,Claude_Code,claude_code_2.1.7,3.10550762,38.01,AppWorld,0.86,gemini,gemini-3-pro,100,0.36,310.550762
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,22.7647676,47.65,AppWorld,0.77,claude,claude-opus-4.5,100,0.68,2276.47676
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0,0,AppWorld,0,gpt52,gpt-5.2,100,0,0
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,8.69552842,33.49,AppWorld,0.98,gemini,gemini-3-pro,100,0.582,869.552842
-smolagent,smolagents,Smolagent,smolagents_1.24.0,5.5851091,41.07,AppWorld,0.82,claude,claude-opus-4.5,100,0.7,558.51091
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.5503114225,51.59,AppWorld,0.61,gpt52,gpt-5.2,100,0.071,55.03114225
-smolagent,smolagents,Smolagent,smolagents_1.24.0,2.54254602,49.13,AppWorld,0.71,gemini,gemini-3-pro,100,0.13,254.254602
-litellm,litellm-react,React,exgentic_0.1.0,11.32465995,21.99,AppWorld,0.83,claude,claude-opus-4.5,100,0.61,1132.465995
-litellm,litellm-react,React,exgentic_0.1.0,0,0,AppWorld,0,gpt52,gpt-5.2,100,0,0
-litellm,litellm-react,React,exgentic_0.1.0,1.88187702,21.76,AppWorld,0.99,gemini,gemini-3-pro,100,0.505,188.187702
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,3.43320175,20.06,AppWorld,0.82,claude,claude-opus-4.5,100,0.64,343.320175
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.3637131575,10.05,AppWorld,1,gpt52,gpt-5.2,100,0.22,36.37131575
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,1.30485292,22.59,AppWorld,1,gemini,gemini-3-pro,100,0.55,130.485292
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,11.65879637,31.03921569,BrowseComp+,0.8431372549,claude,claude-opus-4.5,51,0.5294117647,594.598615
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.4293807,8.97,BrowseComp+,1,gpt52,gpt-5.2,100,0.43,42.93807
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,2.84529442,22.88,BrowseComp+,0.7,gemini,gemini-3-pro,100,0.51,284.529442
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,7.59245735,27.18,BrowseComp+,1,claude,claude-opus-4.5,100,0.61,759.245735
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.38043726,14.27,BrowseComp+,1,gpt52,gpt-5.2,100,0.48,38.043726
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.6433071111,8.454545455,BrowseComp+,0.6060606061,gemini,gemini-3-pro,99,0.3333333333,63.687404
-smolagent,smolagents,Smolagent,smolagents_1.24.0,6.30372505,24.16,BrowseComp+,1,claude,claude-opus-4.5,100,0.61,630.372505
 smolagent,smolagents,Smolagent,smolagents_1.24.0,0.17156755,6.57,BrowseComp+,0.99,gpt52,gpt-5.2,100,0.26,17.156755
-smolagent,smolagents,Smolagent,smolagents_1.24.0,2.38885126,29.63,BrowseComp+,0.69,gemini,gemini-3-pro,100,0.57,238.885126
-litellm,litellm-react,React,exgentic_0.1.0,7.09363285,21.66,BrowseComp+,0.93,claude,claude-opus-4.5,100,0.49,709.363285
-litellm,litellm-react,React,exgentic_0.1.0,0.296172625,8.14,BrowseComp+,0.99,gpt52,gpt-5.2,100,0.46,29.6172625
-litellm,litellm-react,React,exgentic_0.1.0,0.43943356,7.85,BrowseComp+,0.99,gemini,gemini-3-pro,100,0.48,43.943356
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,7.09363285,21.66,BrowseComp+,0.93,claude,claude-opus-4.5,100,0.49,709.363285
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.296172625,8.14,BrowseComp+,0.99,gpt52,gpt-5.2,100,0.46,29.6172625
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.43943356,7.85,BrowseComp+,0.99,gemini,gemini-3-pro,100,0.48,43.943356
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,5.604360979,31.7628866,SWE-bench,1,claude,claude-opus-4.5,97,0.7422680412,543.623015
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.939754725,23.99,SWE-bench,1,gpt52,gpt-5.2,100,0.58,93.9754725
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,3.67966876,43.72,SWE-bench,1,gemini,gemini-3-pro,100,0.67,367.966876
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,2.961261205,34.09638554,SWE-bench,1,claude,claude-opus-4.5,83,0.8072289157,245.78468
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.2590178359,20.44444444,SWE-bench,1,gpt52,gpt-5.2,99,0.5454545455,25.64276575
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,1.579177979,32.36170213,SWE-bench,1,gemini,gemini-3-pro,94,0.7234042553,148.44273
-smolagent,smolagents,Smolagent,smolagents_1.24.0,4.85218325,39.13,SWE-bench,1,claude,claude-opus-4.5,100,0.65,485.218325
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.4503179545,19.97979798,SWE-bench,1,gpt52,gpt-5.2,99,0.5252525253,44.5814775
-smolagent,smolagents,Smolagent,smolagents_1.24.0,2.209661838,38.1010101,SWE-bench,1,gemini,gemini-3-pro,99,0.7575757576,218.756522
-litellm,litellm-react,React,exgentic_0.1.0,3.971294949,43.44444444,SWE-bench,1,claude,claude-opus-4.5,99,0.6060606061,393.1582
-litellm,litellm-react,React,exgentic_0.1.0,0.247620555,20.47,SWE-bench,1,gpt52,gpt-5.2,100,0.57,24.7620555
-litellm,litellm-react,React,exgentic_0.1.0,0.69561342,32.55,SWE-bench,1,gemini,gemini-3-pro,100,0.71,69.561342
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,3.971294949,43.44444444,SWE-bench,1,claude,claude-opus-4.5,99,0.6060606061,393.1582
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.247620555,20.47,SWE-bench,1,gpt52,gpt-5.2,100,0.57,24.7620555
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.69561342,32.55,SWE-bench,1,gemini,gemini-3-pro,100,0.71,69.561342
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,1.2993549,11.5,TauBench-Airline,1,claude,claude-opus-4.5,50,0.66,64.967745
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.21338296,10.18,TauBench-Airline,1,gpt52,gpt-5.2,50,0.48,10.669148
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.3394426,12.62,TauBench-Airline,1,gemini,gemini-3-pro,50,0.7,16.97213
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.7165842,12.22,TauBench-Airline,1,claude,claude-opus-4.5,50,0.74,35.82921
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.10711806,11.4,TauBench-Airline,1,gpt52,gpt-5.2,50,0.5,5.355903
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.21380212,10.9,TauBench-Airline,1,gemini,gemini-3-pro,50,0.62,10.690106
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.7801528,11.88,TauBench-Airline,1,claude,claude-opus-4.5,50,0.72,39.00764
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.293271755,10.68,TauBench-Airline,1,gpt52,gpt-5.2,50,0.6,14.66358775
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.19590488,12.28,TauBench-Airline,1,gemini,gemini-3-pro,50,0.68,9.795244
-litellm,litellm-react,React,exgentic_0.1.0,0.4692376,10,TauBench-Airline,1,claude,claude-opus-4.5,50,0.66,23.46188
-litellm,litellm-react,React,exgentic_0.1.0,0.125104735,11.22,TauBench-Airline,1,gpt52,gpt-5.2,50,0.54,6.25523675
-litellm,litellm-react,React,exgentic_0.1.0,0.15854968,10.14,TauBench-Airline,1,gemini,gemini-3-pro,50,0.7,7.927484
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.4692376,10,TauBench-Airline,1,claude,claude-opus-4.5,50,0.66,23.46188
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.125104735,11.22,TauBench-Airline,1,gpt52,gpt-5.2,50,0.54,6.25523675
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.15854968,10.14,TauBench-Airline,1,gemini,gemini-3-pro,50,0.7,7.927484
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,1.59852325,12.54,TauBench-Retail,1,claude,claude-opus-4.5,100,0.83,159.852325
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.1200643675,9.92,TauBench-Retail,0.98,gpt52,gpt-5.2,100,0.51,12.00643675
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.18525812,11.18,TauBench-Retail,1,gemini,gemini-3-pro,100,0.7804878049,18.525812
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.5495812,12.54,TauBench-Retail,1,claude,claude-opus-4.5,100,0.85,54.95812
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.107783305,9.55,TauBench-Retail,0.99,gpt52,gpt-5.2,100,0.5353535354,10.7783305
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.266485,10.62,TauBench-Retail,1,gemini,gemini-3-pro,100,0.73,26.6485
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.6711861,11.71,TauBench-Retail,1,claude,claude-opus-4.5,100,0.78,67.11861
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.2519468525,11.08,TauBench-Retail,1,gpt52,gpt-5.2,100,0.68,25.19468525
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.2056864,11.3,TauBench-Retail,1,gemini,gemini-3-pro,100,0.7575757576,20.56864
-litellm,litellm-react,React,exgentic_0.1.0,0.46742525,11.33,TauBench-Retail,1,claude,claude-opus-4.5,100,0.78,46.742525
-litellm,litellm-react,React,exgentic_0.1.0,0.1109317125,10.33,TauBench-Retail,1,gpt52,gpt-5.2,100,0.73,11.09317125
-litellm,litellm-react,React,exgentic_0.1.0,0.15649692,11.25,TauBench-Retail,1,gemini,gemini-3-pro,100,0.82,15.649692
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.46742525,11.33,TauBench-Retail,1,claude,claude-opus-4.5,100,0.78,46.742525
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.1109317125,10.33,TauBench-Retail,1,gpt52,gpt-5.2,100,0.73,11.09317125
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.15649692,11.25,TauBench-Retail,1,gemini,gemini-3-pro,100,0.82,15.649692
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,2.4488646,18.71,TauBench-Telecom,1,claude,claude-opus-4.5,100,0.76,244.88646
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.0980149625,9.36,TauBench-Telecom,1,gpt52,gpt-5.2,100,0.55,9.80149625
-claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.21064224,9.9,TauBench-Telecom,1,gemini,gemini-3-pro,100,0.6851851852,21.064224
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,1.2459395,17.15,TauBench-Telecom,1,claude,claude-opus-4.5,100,0.84,124.59395
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.147175455,9.92,TauBench-Telecom,1,gpt52,gpt-5.2,100,0.53,14.7175455
-openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.53513152,10.82,TauBench-Telecom,0.89,gemini,gemini-3-pro,100,0.8876404494,53.513152
-smolagent,smolagents,Smolagent,smolagents_1.24.0,1.0643879,13.77,TauBench-Telecom,1,claude,claude-opus-4.5,100,0.58,106.43879
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.3035534775,10.11,TauBench-Telecom,1,gpt52,gpt-5.2,100,0.71,30.35534775
-smolagent,smolagents,Smolagent,smolagents_1.24.0,0.3461699,12.71,TauBench-Telecom,1,gemini,gemini-3-pro,100,0.88,34.61699
-litellm,litellm-react,React,exgentic_0.1.0,0.91578125,17.22,TauBench-Telecom,1,claude,claude-opus-4.5,100,0.76,91.578125
-litellm,litellm-react,React,exgentic_0.1.0,0.1408951075,10.18,TauBench-Telecom,0.99,gpt52,gpt-5.2,100,0.5353535354,14.08951075
-litellm,litellm-react,React,exgentic_0.1.0,0.29797836,14.84,TauBench-Telecom,1,gemini,gemini-3-pro,100,0.73,29.797836
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.91578125,17.22,TauBench-Telecom,1,claude,claude-opus-4.5,100,0.76,91.578125
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.1408951075,10.18,TauBench-Telecom,0.99,gpt52,gpt-5.2,100,0.5353535354,14.08951075
-litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.29797836,14.84,TauBench-Telecom,1,gemini,gemini-3-pro,100,0.73,29.797836

 agent,agent_normalized,visible_agent_name,agent_version,avg_cost,avg_steps,benchmark,finished_pct,model,model_normalized,num_tasks,score,total_cost
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.5026066430000001,18.94,AppWorld,0.46,deepseek,deepseek-v3.2,100,0.03,50.26066430000001
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.10630119119999998,20.82,BrowseComp+,0.98,deepseek,deepseek-v3.2,100,0.48,10.630119119999998
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.20144125259999995,54.16,SWE-bench,0.89,deepseek,deepseek-v3.2,100,0.64,20.144125259999996
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.028655971600000007,12.06,TauBench-Airline,0.98,deepseek,deepseek-v3.2,50,0.28,1.4327985800000003
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.043596285400000004,12.51,TauBench-Retail,0.99,deepseek,deepseek-v3.2,100,0.65,4.35962854
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.07225669540000003,20.03,TauBench-Telecom,0.97,deepseek,deepseek-v3.2,100,0.61,7.225669540000003
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.8930315160000002,36.62,AppWorld,0.78,kimi,kimi-k2.5,100,0.08,89.30315160000002
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.7320507719999999,25.02,BrowseComp+,0.86,kimi,kimi-k2.5,100,0.56,73.20507719999999
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.7822629151515151,38.484848484848484,SWE-bench,0.7474747474747475,kimi,kimi-k2.5,100,0.5204081632653061,77.4440286
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.032654508000000006,2.58,TauBench-Airline,0.9,kimi,kimi-k2.5,50,0.12,1.6327254000000002
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.018570114000000006,0.56,TauBench-Retail,1.0,kimi,kimi-k2.5,100,0.03,1.8570114000000004
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.005500535999999998,0.0,TauBench-Telecom,1.0,kimi,kimi-k2.5,100,0.0,0.5500535999999998
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.0,0.0,AppWorld,0.0,gpt52,gpt-5.2,100,0.0,0.0
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.4293807000000001,8.97,BrowseComp+,1.0,gpt52,gpt-5.2,100,0.43,42.93807000000001
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.9397547250000001,23.99,SWE-bench,1.0,gpt52,gpt-5.2,100,0.58,93.97547250000001
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.21338296,10.18,TauBench-Airline,1.0,gpt52,gpt-5.2,50,0.48,10.669148
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.1200643675,9.92,TauBench-Retail,0.98,gpt52,gpt-5.2,100,0.64,12.00643675
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.0980149625,9.36,TauBench-Telecom,1.0,gpt52,gpt-5.2,100,0.55,9.80149625
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,13.08379215,49.69,AppWorld,0.74,claude,claude-opus-4.5,100,0.66,1308.3792150000002
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,11.658796372549022,31.03921568627451,BrowseComp+,0.8431372549019608,claude,claude-opus-4.5,51,0.5294117647058824,594.5986150000001
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,5.604360979381441,31.762886597938145,SWE-bench,1.0,claude,claude-opus-4.5,97,0.7422680412371134,543.6230149999998
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,1.2993549000000004,11.5,TauBench-Airline,1.0,claude,claude-opus-4.5,50,0.66,64.96774500000002
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,1.5985232500000006,12.54,TauBench-Retail,1.0,claude,claude-opus-4.5,100,0.83,159.85232500000006
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,2.448864600000001,18.71,TauBench-Telecom,1.0,claude,claude-opus-4.5,100,0.76,244.8864600000001
 claudecode,claude-code,Claude_Code,claude_code_2.1.7,3.10550762,38.01,AppWorld,0.86,gemini,gemini-3-pro,100,0.36,310.550762
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,2.8452944199999997,22.88,BrowseComp+,0.7,gemini,gemini-3-pro,100,0.51,284.52944199999996
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,3.679668759999999,43.72,SWE-bench,1.0,gemini,gemini-3-pro,100,0.67,367.9668759999999
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.3394426000000001,12.62,TauBench-Airline,1.0,gemini,gemini-3-pro,50,0.7,16.972130000000003
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.18525811999999994,11.18,TauBench-Retail,1.0,gemini,gemini-3-pro,100,0.71,18.525811999999995
+claudecode,claude-code,Claude_Code,claude_code_2.1.7,0.21064223999999993,9.9,TauBench-Telecom,1.0,gemini,gemini-3-pro,100,0.71,21.064223999999992
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.3050881588000001,15.63,AppWorld,0.37,deepseek,deepseek-v3.2,100,0.06,30.50881588000001
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.09532471040000004,13.13,BrowseComp+,0.62,deepseek,deepseek-v3.2,100,0.3,9.532471040000004
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.11237704080000001,41.37,SWE-bench,0.38,deepseek,deepseek-v3.2,100,0.7368421052631579,11.23770408
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.016994613600000003,6.72,TauBench-Airline,0.94,deepseek,deepseek-v3.2,50,0.2,0.8497306800000002
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.011356187600000003,4.82,TauBench-Retail,0.97,deepseek,deepseek-v3.2,100,0.19,1.1356187600000003
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.023369379999999995,7.0,TauBench-Telecom,0.97,deepseek,deepseek-v3.2,100,0.18,2.3369379999999995
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.3024222959999999,19.59,AppWorld,0.48,kimi,kimi-k2.5,100,0.08,30.242229599999995
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.25144734599999996,17.61,BrowseComp+,0.53,kimi,kimi-k2.5,100,0.35,25.144734599999996
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.3341831760000001,38.74,SWE-bench,0.78,kimi,kimi-k2.5,100,0.5670103092783505,33.41831760000001
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.004579271999999998,0.5,TauBench-Airline,0.88,kimi,kimi-k2.5,50,0.0,0.2289635999999999
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.0042811740000000004,0.41,TauBench-Retail,0.96,kimi,kimi-k2.5,100,0.01,0.42811740000000004
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.0046396319999999994,0.0,TauBench-Telecom,1.0,kimi,kimi-k2.5,100,0.0,0.46396319999999996
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.0,0.0,AppWorld,0.0,gpt52,gpt-5.2,100,0.0,0.0
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.3804372600000001,14.27,BrowseComp+,1.0,gpt52,gpt-5.2,100,0.48,38.043726000000014
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.259017835858586,20.444444444444443,SWE-bench,1.0,gpt52,gpt-5.2,99,0.5454545454545454,25.642765750000013
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.10711806000000001,11.4,TauBench-Airline,1.0,gpt52,gpt-5.2,50,0.5,5.3559030000000005
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.10778330499999997,9.55,TauBench-Retail,0.99,gpt52,gpt-5.2,100,0.53,10.778330499999997
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.14717545500000007,9.92,TauBench-Telecom,1.0,gpt52,gpt-5.2,100,0.53,14.717545500000005
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,22.764767600000006,47.65,AppWorld,0.77,claude,claude-opus-4.5,100,0.68,2276.4767600000005
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,7.592457349999998,27.18,BrowseComp+,1.0,claude,claude-opus-4.5,100,0.61,759.2457349999999
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,2.9612612048192783,34.096385542168676,SWE-bench,1.0,claude,claude-opus-4.5,83,0.8072289156626506,245.7846800000001
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.7165842000000003,12.22,TauBench-Airline,1.0,claude,claude-opus-4.5,50,0.74,35.82921000000002
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.5495812000000001,12.54,TauBench-Retail,1.0,claude,claude-opus-4.5,100,0.85,54.958120000000015
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,1.2459395,17.15,TauBench-Telecom,1.0,claude,claude-opus-4.5,100,0.84,124.59395
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,8.695528419999997,33.49,AppWorld,0.98,gemini,gemini-3-pro,100,0.57,869.5528419999997
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.6433071111111112,8.454545454545455,BrowseComp+,0.6060606060606061,gemini,gemini-3-pro,99,0.3333333333333333,63.687404
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,1.5791779787234044,32.361702127659576,SWE-bench,1.0,gemini,gemini-3-pro,94,0.723404255319149,148.44273
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.21380211999999996,10.9,TauBench-Airline,1.0,gemini,gemini-3-pro,50,0.62,10.690105999999998
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.26648499999999986,10.62,TauBench-Retail,1.0,gemini,gemini-3-pro,100,0.73,26.648499999999988
+openaimcp,openai-mcp,OpenAI_Solo,openai_sdk_0.7.0,0.5351315199999997,10.82,TauBench-Telecom,0.89,gemini,gemini-3-pro,100,0.79,53.51315199999998
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.2420605992000001,54.69,AppWorld,0.81,deepseek,deepseek-v3.2,100,0.13,24.20605992000001
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.07768438359999996,12.01,BrowseComp+,1.0,deepseek,deepseek-v3.2,100,0.21,7.768438359999997
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.5519223772,69.43,SWE-bench,0.98,deepseek,deepseek-v3.2,100,0.56,55.19223772000001
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.055420856,18.94,TauBench-Airline,1.0,deepseek,deepseek-v3.2,50,0.6,2.7710428
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.05129110320000001,14.06,TauBench-Retail,0.98,deepseek,deepseek-v3.2,100,0.77,5.1291103200000006
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.06806756960000002,18.3,TauBench-Telecom,1.0,deepseek,deepseek-v3.2,100,0.84,6.806756960000001
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.6740286120000001,50.63,AppWorld,0.91,kimi,kimi-k2.5,100,0.11,67.4028612
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.4948784999999996,29.79,BrowseComp+,0.95,kimi,kimi-k2.5,100,0.33,49.48784999999996
+smolagent,smolagents,Smolagent,smolagents_1.24.0,1.3741787032258057,65.91397849462365,SWE-bench,0.8709677419354839,kimi,kimi-k2.5,100,0.5760869565217391,127.79861939999994
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.09928738799999998,13.0,TauBench-Airline,1.0,kimi,kimi-k2.5,50,0.56,4.964369399999999
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.06753205714285712,12.887755102040817,TauBench-Retail,1.0,kimi,kimi-k2.5,100,0.7244897959183674,6.618141599999998
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.15341341818181825,19.141414141414142,TauBench-Telecom,1.0,kimi,kimi-k2.5,100,0.7070707070707071,15.187928400000008
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.5503114225,51.59,AppWorld,0.61,gpt52,gpt-5.2,100,0.07,55.03114225
 smolagent,smolagents,Smolagent,smolagents_1.24.0,0.17156755,6.57,BrowseComp+,0.99,gpt52,gpt-5.2,100,0.26,17.156755
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.45031795454545454,19.97979797979798,SWE-bench,1.0,gpt52,gpt-5.2,99,0.5252525252525253,44.5814775
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.2932717550000001,10.68,TauBench-Airline,1.0,gpt52,gpt-5.2,50,0.6,14.663587750000003
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.25194685249999993,11.08,TauBench-Retail,1.0,gpt52,gpt-5.2,100,0.68,25.194685249999992
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.30355347749999995,10.11,TauBench-Telecom,1.0,gpt52,gpt-5.2,100,0.71,30.355347749999993
+smolagent,smolagents,Smolagent,smolagents_1.24.0,5.585109100000002,41.07,AppWorld,0.82,claude,claude-opus-4.5,100,0.7,558.5109100000002
+smolagent,smolagents,Smolagent,smolagents_1.24.0,6.303725049999997,24.16,BrowseComp+,1.0,claude,claude-opus-4.5,100,0.61,630.3725049999997
+smolagent,smolagents,Smolagent,smolagents_1.24.0,4.8521832499999995,39.13,SWE-bench,1.0,claude,claude-opus-4.5,100,0.65,485.218325
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.7801527999999999,11.88,TauBench-Airline,1.0,claude,claude-opus-4.5,50,0.72,39.007639999999995
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.6711861,11.71,TauBench-Retail,1.0,claude,claude-opus-4.5,100,0.78,67.11861
+smolagent,smolagents,Smolagent,smolagents_1.24.0,1.0643879000000003,13.77,TauBench-Telecom,1.0,claude,claude-opus-4.5,100,0.58,106.43879000000003
+smolagent,smolagents,Smolagent,smolagents_1.24.0,2.542546020000001,49.13,AppWorld,0.71,gemini,gemini-3-pro,100,0.13,254.2546020000001
+smolagent,smolagents,Smolagent,smolagents_1.24.0,2.388851260000001,29.63,BrowseComp+,0.69,gemini,gemini-3-pro,100,0.57,238.88512600000007
+smolagent,smolagents,Smolagent,smolagents_1.24.0,2.209661838383839,38.101010101010104,SWE-bench,1.0,gemini,gemini-3-pro,99,0.7575757575757576,218.75652200000005
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.19590487999999995,12.28,TauBench-Airline,1.0,gemini,gemini-3-pro,50,0.68,9.795243999999997
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.2056864,11.3,TauBench-Retail,1.0,gemini,gemini-3-pro,100,0.75,20.56864
+smolagent,smolagents,Smolagent,smolagents_1.24.0,0.3461698999999998,12.71,TauBench-Telecom,1.0,gemini,gemini-3-pro,100,0.88,34.61698999999998
+litellm,litellm-react,React,exgentic_0.1.0,0.8332616264000001,31.69,AppWorld,0.88,deepseek,deepseek-v3.2,100,0.09,83.32616264
+litellm,litellm-react,React,exgentic_0.1.0,0.1679167768,34.19,BrowseComp+,0.98,deepseek,deepseek-v3.2,100,0.36,16.79167768
+litellm,litellm-react,React,exgentic_0.1.0,0.47889091680000007,77.44,SWE-bench,0.94,deepseek,deepseek-v3.2,100,0.6875,47.88909168000001
+litellm,litellm-react,React,exgentic_0.1.0,0.033952516,15.28,TauBench-Airline,1.0,deepseek,deepseek-v3.2,50,0.56,1.6976258
+litellm,litellm-react,React,exgentic_0.1.0,0.029598555599999986,14.44,TauBench-Retail,1.0,deepseek,deepseek-v3.2,100,0.82,2.9598555599999985
+litellm,litellm-react,React,exgentic_0.1.0,0.04940463720000001,18.11,TauBench-Telecom,1.0,deepseek,deepseek-v3.2,100,0.71,4.940463720000001
+litellm,litellm-react,React,exgentic_0.1.0,0.8925547499999997,25.05,AppWorld,0.9,kimi,kimi-k2.5,100,0.09,89.25547499999998
+litellm,litellm-react,React,exgentic_0.1.0,0.3791127780000001,26.27,BrowseComp+,0.74,kimi,kimi-k2.5,100,0.34,37.91127780000001
+litellm,litellm-react,React,exgentic_0.1.0,0.590102075510204,48.765306122448976,SWE-bench,0.8775510204081632,kimi,kimi-k2.5,100,0.5714285714285714,57.830003399999995
+litellm,litellm-react,React,exgentic_0.1.0,0.03922161599999999,9.6,TauBench-Airline,1.0,kimi,kimi-k2.5,50,0.62,1.9610807999999995
+litellm,litellm-react,React,exgentic_0.1.0,0.04560821212121211,12.313131313131313,TauBench-Retail,1.0,kimi,kimi-k2.5,100,0.6464646464646465,4.515212999999999
+litellm,litellm-react,React,exgentic_0.1.0,0.092763078,16.14,TauBench-Telecom,1.0,kimi,kimi-k2.5,100,0.83,9.2763078
+litellm,litellm-react,React,exgentic_0.1.0,0.0,0.0,AppWorld,0.0,gpt52,gpt-5.2,100,0.0,0.0
+litellm,litellm-react,React,exgentic_0.1.0,0.2961726250000001,8.14,BrowseComp+,0.99,gpt52,gpt-5.2,100,0.46,29.617262500000006
+litellm,litellm-react,React,exgentic_0.1.0,0.247620555,20.47,SWE-bench,1.0,gpt52,gpt-5.2,100,0.57,24.7620555
+litellm,litellm-react,React,exgentic_0.1.0,0.12510473500000002,11.22,TauBench-Airline,1.0,gpt52,gpt-5.2,50,0.54,6.255236750000001
+litellm,litellm-react,React,exgentic_0.1.0,0.11093171249999992,10.33,TauBench-Retail,1.0,gpt52,gpt-5.2,100,0.73,11.093171249999992
+litellm,litellm-react,React,exgentic_0.1.0,0.14089510749999992,10.18,TauBench-Telecom,0.99,gpt52,gpt-5.2,100,0.53,14.089510749999993
+litellm,litellm-react,React,exgentic_0.1.0,11.324659950000004,21.99,AppWorld,0.83,claude,claude-opus-4.5,100,0.61,1132.4659950000005
+litellm,litellm-react,React,exgentic_0.1.0,7.093632849999999,21.66,BrowseComp+,0.93,claude,claude-opus-4.5,100,0.49,709.3632849999999
+litellm,litellm-react,React,exgentic_0.1.0,3.971294949494951,43.44444444444444,SWE-bench,1.0,claude,claude-opus-4.5,99,0.6060606060606061,393.15820000000014
+litellm,litellm-react,React,exgentic_0.1.0,0.46923760000000003,10.0,TauBench-Airline,1.0,claude,claude-opus-4.5,50,0.66,23.46188
+litellm,litellm-react,React,exgentic_0.1.0,0.46742524999999985,11.33,TauBench-Retail,1.0,claude,claude-opus-4.5,100,0.78,46.742524999999986
+litellm,litellm-react,React,exgentic_0.1.0,0.9157812500000004,17.22,TauBench-Telecom,1.0,claude,claude-opus-4.5,100,0.76,91.57812500000004
+litellm,litellm-react,React,exgentic_0.1.0,1.8818770200000003,21.76,AppWorld,0.99,gemini,gemini-3-pro,100,0.5,188.18770200000003
+litellm,litellm-react,React,exgentic_0.1.0,0.4394335600000001,7.85,BrowseComp+,0.99,gemini,gemini-3-pro,100,0.48,43.94335600000001
+litellm,litellm-react,React,exgentic_0.1.0,0.6956134199999998,32.55,SWE-bench,1.0,gemini,gemini-3-pro,100,0.71,69.56134199999998
+litellm,litellm-react,React,exgentic_0.1.0,0.15854967999999997,10.14,TauBench-Airline,1.0,gemini,gemini-3-pro,50,0.7,7.927483999999998
+litellm,litellm-react,React,exgentic_0.1.0,0.15649691999999996,11.25,TauBench-Retail,1.0,gemini,gemini-3-pro,100,0.82,15.649691999999995
+litellm,litellm-react,React,exgentic_0.1.0,0.29797836000000005,14.84,TauBench-Telecom,1.0,gemini,gemini-3-pro,100,0.73,29.797836000000004
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.23222570840000004,28.51,AppWorld,0.52,deepseek,deepseek-v3.2,100,0.04,23.222570840000003
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.1679167768,34.19,BrowseComp+,0.98,deepseek,deepseek-v3.2,100,0.36,16.79167768
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.47889091680000007,77.44,SWE-bench,0.94,deepseek,deepseek-v3.2,100,0.6875,47.88909168000001
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.033952516,15.28,TauBench-Airline,1.0,deepseek,deepseek-v3.2,50,0.56,1.6976258
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.029598555599999986,14.44,TauBench-Retail,1.0,deepseek,deepseek-v3.2,100,0.82,2.9598555599999985
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.04940463720000001,18.11,TauBench-Telecom,1.0,deepseek,deepseek-v3.2,100,0.71,4.940463720000001
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.6031619339999997,23.53,AppWorld,0.89,kimi,kimi-k2.5,100,0.1,60.316193399999975
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.3791127780000001,26.27,BrowseComp+,0.74,kimi,kimi-k2.5,100,0.34,37.91127780000001
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.590102075510204,48.765306122448976,SWE-bench,0.8775510204081632,kimi,kimi-k2.5,100,0.5714285714285714,57.830003399999995
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.03922161599999999,9.6,TauBench-Airline,1.0,kimi,kimi-k2.5,50,0.62,1.9610807999999995
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.04560821212121211,12.313131313131313,TauBench-Retail,1.0,kimi,kimi-k2.5,100,0.6464646464646465,4.515212999999999
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.092763078,16.14,TauBench-Telecom,1.0,kimi,kimi-k2.5,100,0.83,9.2763078
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.36371315749999994,10.05,AppWorld,1.0,gpt52,gpt-5.2,100,0.22,36.371315749999994
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.2961726250000001,8.14,BrowseComp+,0.99,gpt52,gpt-5.2,100,0.46,29.617262500000006
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.247620555,20.47,SWE-bench,1.0,gpt52,gpt-5.2,100,0.57,24.7620555
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.12510473500000002,11.22,TauBench-Airline,1.0,gpt52,gpt-5.2,50,0.54,6.255236750000001
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.11093171249999992,10.33,TauBench-Retail,1.0,gpt52,gpt-5.2,100,0.73,11.093171249999992
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.14089510749999992,10.18,TauBench-Telecom,0.99,gpt52,gpt-5.2,100,0.53,14.089510749999993
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,3.4332017499999994,20.06,AppWorld,0.82,claude,claude-opus-4.5,100,0.64,343.32017499999995
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,7.093632849999999,21.66,BrowseComp+,0.93,claude,claude-opus-4.5,100,0.49,709.3632849999999
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,3.971294949494951,43.44444444444444,SWE-bench,1.0,claude,claude-opus-4.5,99,0.6060606060606061,393.15820000000014
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.46923760000000003,10.0,TauBench-Airline,1.0,claude,claude-opus-4.5,50,0.66,23.46188
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.46742524999999985,11.33,TauBench-Retail,1.0,claude,claude-opus-4.5,100,0.78,46.742524999999986
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.9157812500000004,17.22,TauBench-Telecom,1.0,claude,claude-opus-4.5,100,0.76,91.57812500000004
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,1.30485292,22.59,AppWorld,1.0,gemini,gemini-3-pro,100,0.55,130.48529200000002
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.4394335600000001,7.85,BrowseComp+,0.99,gemini,gemini-3-pro,100,0.48,43.94335600000001
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.6956134199999998,32.55,SWE-bench,1.0,gemini,gemini-3-pro,100,0.71,69.56134199999998
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.15854967999999997,10.14,TauBench-Airline,1.0,gemini,gemini-3-pro,50,0.7,7.927483999999998
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.15649691999999996,11.25,TauBench-Retail,1.0,gemini,gemini-3-pro,100,0.82,15.649691999999995
+litellm-shortlist,litellm-shortlist,React_+_Shortlisting,exgentic_0.1.0_�_litellm_1.79.1,0.29797836000000005,14.84,TauBench-Telecom,1.0,gemini,gemini-3-pro,100,0.73,29.797836000000004

results.csv.timestamp CHANGED Viewed

	@@ -1 +1 @@
1	- ~~Mon~~ ~~Mar~~ 30 14:50:35 ~~UTC 2026~~


1	+ Wednesday, May 13, 2026 12:25:52 PM

scripts/leaderboard.js CHANGED Viewed

@@ -1,8 +1,8 @@
 // ===== 9. LEADERBOARD =====
 const BENCHMARKS=['AppWorld','BrowseComp+','SWE-bench','TauBench-Airline','TauBench-Retail','TauBench-Telecom'];
 const BENCH_SHORT={'AppWorld':'App','BrowseComp+':'Browse','SWE-bench':'SWE','TauBench-Airline':'Tau-Air','TauBench-Retail':'Tau-Ret','TauBench-Telecom':'Tau-Tel'};
-const MODEL_DISPLAY={'claude-opus-4.5':'Claude Opus 4.5','gpt-5.2':'GPT 5.2','gemini-3-pro':'Gemini Pro 3'};
-const MODEL_URLS={'claude-opus-4.5':'https://www.anthropic.com/claude','gpt-5.2':'https://openai.com/','gemini-3-pro':'https://deepmind.google/technologies/gemini/'};
 const AGENT_DISPLAY={'Claude_Code':'Claude Code','OpenAI_Solo':'OpenAI Solo','Smolagent':'Smolagent','React':'React','React_+_Shortlisting':'React + Shortlist'};
 const AGENT_URLS={'Claude_Code':'https://github.com/anthropics/claude-code','OpenAI_Solo':'https://github.com/openai/openai-agents-python','Smolagent':'https://github.com/huggingface/smolagents','React':'https://github.com/BerriAI/litellm','React_+_Shortlisting':'https://github.com/BerriAI/litellm'};
@@ -27,10 +27,12 @@ function processData(rows,modelFilter){
   const BENCH_WEIGHT={};
   BENCHMARKS.forEach(b=>{BENCH_WEIGHT[b]=b.startsWith('TauBench')?1/12:1/4});
   return Object.values(groups).map(g=>{
     const bs=BENCHMARKS.map(b=>g.benchmarks[b]||0);
-    let wSum=0,wTotal=0;
-    BENCHMARKS.forEach((b,i)=>{if(bs[i]>0){wSum+=bs[i]*BENCH_WEIGHT[b];wTotal+=BENCH_WEIGHT[b]}});
-    const avg=wTotal?wSum/wTotal:0;
     const cs=Object.values(g.costs).filter(c=>c>0);
     const avgCost=cs.length?cs.reduce((a,b)=>a+b,0)/cs.length:0;
     return{...g,avg,avgCost,benchScores:bs};
@@ -81,8 +83,7 @@ function renderTable(data){
     html+=`<td class="score-cell ${scoreClass(row.avg)}"><div class="bar bar-cyan" style="width:${row.avg*100}%"></div><span class="val">${fmtPct(row.avg)}</span></td>`;
     html+=`<td class="cost-cell">$${row.avgCost.toFixed(2)}</td>`;
     row.benchScores.forEach(s=>{
-      const barW=s>0?s*100:0;
-      html+=`<td class="score-cell ${scoreClass(s)}"><div class="bar bar-purple" style="width:${barW}%"></div><span class="val">${s>0?fmtPct(s):'&mdash;'}</span></td>`;
     });
     html+=`</tr>`;return html;
   }).join('');
@@ -98,6 +99,25 @@ function renderTable(data){
       renderTable(data);
     });
   });
 }
 // ===== CHART =====

 // ===== 9. LEADERBOARD =====
 const BENCHMARKS=['AppWorld','BrowseComp+','SWE-bench','TauBench-Airline','TauBench-Retail','TauBench-Telecom'];
 const BENCH_SHORT={'AppWorld':'App','BrowseComp+':'Browse','SWE-bench':'SWE','TauBench-Airline':'Tau-Air','TauBench-Retail':'Tau-Ret','TauBench-Telecom':'Tau-Tel'};
+const MODEL_DISPLAY={'claude-opus-4.5':'Claude Opus 4.5','gpt-5.2':'GPT 5.2','gemini-3-pro':'Gemini Pro 3','deepseek-v3.2':'DeepSeek V3.2','kimi-k2.5':'Kimi K2.5'};
+const MODEL_URLS={'claude-opus-4.5':'https://www.anthropic.com/claude','gpt-5.2':'https://openai.com/','gemini-3-pro':'https://deepmind.google/technologies/gemini/','deepseek-v3.2':'https://www.deepseek.com/','kimi-k2.5':'https://www.moonshot.ai/'};
 const AGENT_DISPLAY={'Claude_Code':'Claude Code','OpenAI_Solo':'OpenAI Solo','Smolagent':'Smolagent','React':'React','React_+_Shortlisting':'React + Shortlist'};
 const AGENT_URLS={'Claude_Code':'https://github.com/anthropics/claude-code','OpenAI_Solo':'https://github.com/openai/openai-agents-python','Smolagent':'https://github.com/huggingface/smolagents','React':'https://github.com/BerriAI/litellm','React_+_Shortlisting':'https://github.com/BerriAI/litellm'};
   const BENCH_WEIGHT={};
   BENCHMARKS.forEach(b=>{BENCH_WEIGHT[b]=b.startsWith('TauBench')?1/12:1/4});
   return Object.values(groups).map(g=>{
+    // Every (model, agent, benchmark) cell is populated; zero is a real score
+    // (TauBench protocol failures, AppWorld tool-limit failures). Include all six.
     const bs=BENCHMARKS.map(b=>g.benchmarks[b]||0);
+    let wSum=0;
+    BENCHMARKS.forEach((b,i)=>{wSum+=bs[i]*BENCH_WEIGHT[b]});
+    const avg=wSum;  // weights sum to 1
     const cs=Object.values(g.costs).filter(c=>c>0);
     const avgCost=cs.length?cs.reduce((a,b)=>a+b,0)/cs.length:0;
     return{...g,avg,avgCost,benchScores:bs};
     html+=`<td class="score-cell ${scoreClass(row.avg)}"><div class="bar bar-cyan" style="width:${row.avg*100}%"></div><span class="val">${fmtPct(row.avg)}</span></td>`;
     html+=`<td class="cost-cell">$${row.avgCost.toFixed(2)}</td>`;
     row.benchScores.forEach(s=>{
+      html+=`<td class="score-cell ${scoreClass(s)}"><div class="bar bar-purple" style="width:${s*100}%"></div><span class="val">${fmtPct(s)}</span></td>`;
     });
     html+=`</tr>`;return html;
   }).join('');
       renderTable(data);
     });
   });
+  // Remove any legacy "See all" button if it exists from a previous render
+  const oldBtn=document.getElementById('lbSeeAll');
+  if(oldBtn)oldBtn.remove();
+  // Toggle bottom-fade hint based on whether the table can still scroll down
+  const tableWrap=document.querySelector('.table-wrap');
+  const updateFade=()=>{
+    const atBottom=tableWrap.scrollTop+tableWrap.clientHeight>=tableWrap.scrollHeight-2;
+    const overflows=tableWrap.scrollHeight>tableWrap.clientHeight+2;
+    tableWrap.classList.toggle('lb-can-scroll',overflows&&!atBottom);
+  };
+  if(!tableWrap.dataset.scrollBound){
+    tableWrap.addEventListener('scroll',updateFade,{passive:true});
+    window.addEventListener('resize',updateFade);
+    tableWrap.dataset.scrollBound='1';
+  }
+  // Defer to next frame so layout is settled
+  requestAnimationFrame(updateFade);
 }
 // ===== CHART =====

scripts/pareto.js DELETED Viewed

@@ -1,386 +0,0 @@
-// SVG Pareto Chart — reads processed data from leaderboard.js via window._chartData
-(() => {
-  const MODEL_DISPLAY = {
-    'claude-opus-4.5': 'Claude Opus 4.5',
-    'gpt-5.2': 'GPT 5.2',
-    'gemini-3-pro': 'Gemini Pro 3'
-  };
-  const MODEL_COLORS = {
-    'claude-opus-4.5': '#5ba8a0',
-    'gpt-5.2': '#9b8ec4',
-    'gemini-3-pro': '#6a8cbe'
-  };
-  const AGENT_DISPLAY = {
-    'Claude_Code': 'Claude Code',
-    'OpenAI_Solo': 'OpenAI Solo',
-    'Smolagent': 'Smolagent',
-    'React': 'React',
-    'React_+_Shortlisting': 'React + Shortlist'
-  };
-  const ns = 'http://www.w3.org/2000/svg';
-  function el(tag, attrs) {
-    const e = document.createElementNS(ns, tag);
-    if (attrs) for (const [k, v] of Object.entries(attrs)) e.setAttribute(k, v);
-    return e;
-  }
-  function makeShape(agent, x, y, color) {
-    const g = el('g');
-    const common = { fill: color, 'fill-opacity': '0.82', stroke: 'rgba(128,128,128,0.3)', 'stroke-width': '1.5' };
-    let shape;
-    if (agent === 'Claude_Code') {
-      const half = 9;
-      const pts = [];
-      for (let i = 0; i < 10; i++) {
-        const r = i % 2 === 0 ? half : half * 0.42;
-        const a = i * Math.PI / 5 - Math.PI / 2;
-        pts.push((x + r * Math.cos(a)).toFixed(1) + ',' + (y + r * Math.sin(a)).toFixed(1));
-      }
-      shape = el('polygon', { points: pts.join(' '), ...common });
-    } else if (agent === 'OpenAI_Solo') {
-      const half = 8;
-      shape = el('polygon', { points: `${x},${y - half} ${x + half},${y} ${x},${y + half} ${x - half},${y}`, ...common });
-    } else if (agent === 'Smolagent') {
-      const half = 8.5;
-      shape = el('polygon', { points: `${x},${y - half} ${x + half},${y + half} ${x - half},${y + half}`, ...common });
-    } else if (agent === 'React_+_Shortlisting') {
-      const half = 6.5;
-      shape = el('rect', { x: x - half, y: y - half, width: half * 2, height: half * 2, rx: 2, ...common });
-    } else {
-      const half = 7;
-      shape = el('circle', { cx: x, cy: y, r: half, ...common });
-    }
-    g.appendChild(shape);
-    // Invisible larger hit area for hover
-    g.appendChild(el('circle', { cx: x, cy: y, r: 20, fill: 'transparent', cursor: 'pointer' }));
-    return g;
-  }
-  function makeLegendShapeSVG(agent) {
-    const svg = document.createElementNS(ns, 'svg');
-    svg.setAttribute('width', '14');
-    svg.setAttribute('height', '14');
-    svg.setAttribute('viewBox', '0 0 14 14');
-    const common = { fill: 'var(--text-muted)', 'fill-opacity': '0.5', stroke: 'var(--text-muted)', 'stroke-width': '0.8' };
-    let shape;
-    if (agent === 'Claude_Code') {
-      const pts = [];
-      for (let i = 0; i < 10; i++) {
-        const r = i % 2 === 0 ? 6 : 6 * 0.42;
-        const a = i * Math.PI / 5 - Math.PI / 2;
-        pts.push((7 + r * Math.cos(a)).toFixed(1) + ',' + (7 + r * Math.sin(a)).toFixed(1));
-      }
-      shape = el('polygon', { points: pts.join(' '), ...common });
-    } else if (agent === 'OpenAI_Solo') {
-      shape = el('polygon', { points: '7,1 11,7 7,13 3,7', ...common });
-    } else if (agent === 'Smolagent') {
-      shape = el('polygon', { points: '7,2 12,12 2,12', ...common });
-    } else if (agent === 'React_+_Shortlisting') {
-      shape = el('rect', { x: 2, y: 2, width: 10, height: 10, rx: 1.5, ...common });
-    } else {
-      shape = el('circle', { cx: 7, cy: 7, r: 5, ...common });
-    }
-    svg.appendChild(shape);
-    return svg;
-  }
-  function renderParetoChart(data) {
-    const container = document.getElementById('paretoChartContainer');
-    if (!container) return;
-    container.innerHTML = '';
-    // Build chart data from processed leaderboard data
-    const points = data.map(d => ({
-      agent: d.agent,
-      model: d.model,
-      cost: d.avgCost,
-      success: d.avg
-    })).filter(d => d.cost > 0 && d.success > 0);
-    if (points.length === 0) return;
-    // Determine unique agents and models
-    const agents = [...new Set(points.map(p => p.agent))];
-    const models = [...new Set(points.map(p => p.model))];
-    // Determine axis ranges with padding
-    const costs = points.map(p => p.cost);
-    const successes = points.map(p => p.success);
-    const xMax = Math.ceil(Math.max(...costs) + 0.5);
-    const xMin = 0;
-    const yMinRaw = Math.min(...successes);
-    const yMaxRaw = Math.max(...successes);
-    const yMin = Math.floor(yMinRaw * 10) / 10 - 0.05;
-    const yMax = Math.ceil(yMaxRaw * 10) / 10 + 0.05;
-    // SVG dimensions
-    const W = 740, H = 420;
-    const ml = 62, mr = 20, mt = 14, mb = 52;
-    const pw = W - ml - mr, ph = H - mt - mb;
-    const sx = c => ml + ((c - xMin) / (xMax - xMin)) * pw;
-    const sy = s => mt + (1 - ((s - yMin) / (yMax - yMin))) * ph;
-    // Pareto frontier
-    const sorted = [...points].sort((a, b) => a.cost - b.cost);
-    const pareto = [];
-    let bestS = -Infinity;
-    for (const p of sorted) {
-      if (p.success > bestS) { pareto.push(p); bestS = p.success; }
-    }
-    // Build DOM
-    const root = document.createElement('div');
-    root.className = 'pareto-embed';
-    const chartWrap = document.createElement('div');
-    chartWrap.className = 'chart-wrap';
-    const svg = document.createElementNS(ns, 'svg');
-    svg.setAttribute('class', 'chart');
-    svg.setAttribute('viewBox', `0 0 ${W} ${H}`);
-    svg.setAttribute('aria-label', 'Pareto frontier: cost vs success rate for agent configurations');
-    // Gradient for pareto area
-    const defs = el('defs');
-    const grad = el('linearGradient', { id: 'pareto-fade-main', x1: '0', y1: '0', x2: '0', y2: '1' });
-    const stop1 = el('stop', { offset: '0%', 'stop-color': '#5ba8a0', 'stop-opacity': '0.08' });
-    const stop2 = el('stop', { offset: '100%', 'stop-color': '#5ba8a0', 'stop-opacity': '0' });
-    grad.appendChild(stop1);
-    grad.appendChild(stop2);
-    defs.appendChild(grad);
-    svg.appendChild(defs);
-    // Grid lines and ticks
-    const xStep = xMax <= 10 ? 1 : 2;
-    const xTicks = [];
-    for (let x = 0; x <= xMax; x += xStep) xTicks.push(x);
-    const yTicks = [];
-    for (let y = Math.ceil(yMin * 10) / 10; y <= yMax; y = Math.round((y + 0.1) * 10) / 10) yTicks.push(y);
-    for (const x of xTicks) {
-      const px = sx(x);
-      svg.appendChild(el('line', { x1: px, y1: mt, x2: px, y2: mt + ph, stroke: 'var(--text-muted)', 'stroke-opacity': '0.15', 'stroke-width': 1 }));
-      const t = el('text', { x: px, y: mt + ph + 22, 'font-size': 11, fill: 'var(--text-muted)', 'text-anchor': 'middle', 'font-family': 'Inter, system-ui', 'font-weight': 400 });
-      t.textContent = '$' + x;
-      svg.appendChild(t);
-    }
-    for (const y of yTicks) {
-      const py = sy(y);
-      svg.appendChild(el('line', { x1: ml, y1: py, x2: ml + pw, y2: py, stroke: 'var(--text-muted)', 'stroke-opacity': '0.15', 'stroke-width': 1 }));
-      const t = el('text', { x: ml - 10, y: py + 4, 'font-size': 11, fill: 'var(--text-muted)', 'text-anchor': 'end', 'font-family': 'Inter, system-ui', 'font-weight': 400 });
-      t.textContent = Math.round(y * 100) + '%';
-      svg.appendChild(t);
-    }
-    // Axes
-    svg.appendChild(el('line', { x1: ml, y1: mt + ph, x2: ml + pw, y2: mt + ph, stroke: 'var(--text-muted)', 'stroke-opacity': '0.25', 'stroke-width': 1 }));
-    svg.appendChild(el('line', { x1: ml, y1: mt, x2: ml, y2: mt + ph, stroke: 'var(--text-muted)', 'stroke-opacity': '0.25', 'stroke-width': 1 }));
-    // Axis labels
-    const xLabel = el('text', { x: ml + pw / 2, y: H - 8, 'text-anchor': 'middle', 'font-size': 12, fill: 'var(--text-muted)', 'font-family': 'Inter, system-ui', 'font-weight': 500 });
-    xLabel.textContent = 'Average cost per task (USD)';
-    svg.appendChild(xLabel);
-    const yLabel = el('text', { x: 14, y: mt + ph / 2, 'text-anchor': 'middle', 'font-size': 12, fill: 'var(--text-muted)', 'font-family': 'Inter, system-ui', 'font-weight': 500, transform: 'rotate(-90 14 ' + (mt + ph / 2) + ')' });
-    yLabel.textContent = 'Success rate';
-    svg.appendChild(yLabel);
-    // Pareto area fill
-    if (pareto.length > 1) {
-      let areaD = 'M ' + sx(pareto[0].cost) + ' ' + sy(pareto[0].success);
-      for (let i = 1; i < pareto.length; i++) areaD += ' L ' + sx(pareto[i].cost) + ' ' + sy(pareto[i].success);
-      areaD += ' L ' + sx(pareto[pareto.length - 1].cost) + ' ' + (mt + ph);
-      areaD += ' L ' + sx(pareto[0].cost) + ' ' + (mt + ph) + ' Z';
-      svg.appendChild(el('path', { d: areaD, fill: 'url(#pareto-fade-main)' }));
-    }
-    // Pareto line
-    const paretoD = pareto.map((p, i) => (i === 0 ? 'M' : 'L') + ' ' + sx(p.cost) + ' ' + sy(p.success)).join(' ');
-    const paretoLine = el('path', { d: paretoD, fill: 'none', stroke: 'var(--text-muted)', 'stroke-opacity': '0.4', 'stroke-width': 1.6, 'stroke-dasharray': '6 4' });
-    svg.appendChild(paretoLine);
-    // Pareto label
-    const lastPareto = pareto[pareto.length - 1];
-    const paretoLabel = el('text', { x: sx(lastPareto.cost) + 8, y: sy(lastPareto.success) - 10, 'font-size': 10, fill: 'var(--text-muted)', 'font-family': 'Inter, system-ui', 'font-style': 'italic' });
-    paretoLabel.textContent = 'Pareto frontier';
-    svg.appendChild(paretoLabel);
-    // Data points
-    const pointEls = [];
-    points.forEach((d, i) => {
-      const x = sx(d.cost);
-      const y = sy(d.success);
-      const color = MODEL_COLORS[d.model] || '#7A7A7A';
-      const g = makeShape(d.agent, x, y, color);
-      g.classList.add('chart-point');
-      g.dataset.agent = d.agent;
-      g.dataset.model = d.model;
-      g.dataset.idx = i;
-      svg.appendChild(g);
-      pointEls.push({ el: g, data: d, x, y });
-    });
-    chartWrap.appendChild(svg);
-    // Tooltip
-    const tooltip = document.createElement('div');
-    tooltip.className = 'tooltip';
-    chartWrap.appendChild(tooltip);
-    root.appendChild(chartWrap);
-    // Legend
-    const legendRow = document.createElement('div');
-    legendRow.className = 'legend-row';
-    // Agents legend group
-    const agentGroup = document.createElement('div');
-    agentGroup.className = 'legend-group';
-    const agentLabel = document.createElement('span');
-    agentLabel.className = 'legend-group-label';
-    agentLabel.textContent = 'Agents';
-    agentGroup.appendChild(agentLabel);
-    agents.forEach(agent => {
-      const item = document.createElement('div');
-      item.className = 'legend-item';
-      item.dataset.filter = 'agent';
-      item.dataset.value = agent;
-      item.appendChild(makeLegendShapeSVG(agent));
-      const span = document.createElement('span');
-      span.textContent = AGENT_DISPLAY[agent] || agent.replace(/_/g, ' ');
-      item.appendChild(span);
-      agentGroup.appendChild(item);
-    });
-    // Models legend group
-    const modelGroup = document.createElement('div');
-    modelGroup.className = 'legend-group';
-    const modelLabel = document.createElement('span');
-    modelLabel.className = 'legend-group-label';
-    modelLabel.textContent = 'Models';
-    modelGroup.appendChild(modelLabel);
-    models.forEach(model => {
-      const item = document.createElement('div');
-      item.className = 'legend-item';
-      item.dataset.filter = 'model';
-      item.dataset.value = model;
-      const dot = document.createElement('span');
-      dot.style.cssText = 'display:inline-block;width:11px;height:11px;border-radius:50%;background:' + (MODEL_COLORS[model] || '#666') + ';border:1px solid rgba(128,128,128,0.2);';
-      item.appendChild(dot);
-      const span = document.createElement('span');
-      span.textContent = MODEL_DISPLAY[model] || model;
-      item.appendChild(span);
-      modelGroup.appendChild(item);
-    });
-    legendRow.appendChild(agentGroup);
-    legendRow.appendChild(modelGroup);
-    root.appendChild(legendRow);
-    container.appendChild(root);
-    // Interaction
-    const activeFilters = { agent: null, model: null };
-    const legendItems = root.querySelectorAll('.legend-item[data-filter]');
-    function applyFilters() {
-      const hasFilter = activeFilters.agent || activeFilters.model;
-      legendItems.forEach(li => {
-        const t = li.dataset.filter;
-        const v = li.dataset.value;
-        if (activeFilters[t] === v) {
-          li.classList.add('active');
-          li.classList.remove('dimmed');
-        } else if (activeFilters[t] && activeFilters[t] !== v) {
-          li.classList.remove('active');
-          li.classList.add('dimmed');
-        } else {
-          li.classList.remove('active');
-          li.classList.remove('dimmed');
-        }
-      });
-      pointEls.forEach(({ el: g, data: d }) => {
-        const matchAgent = !activeFilters.agent || d.agent === activeFilters.agent;
-        const matchModel = !activeFilters.model || d.model === activeFilters.model;
-        const visible = !hasFilter || (matchAgent && matchModel);
-        g.style.opacity = visible ? '1' : '0.08';
-        g.style.filter = 'none';
-      });
-      paretoLine.style.opacity = hasFilter ? '0.1' : '1';
-    }
-    function showTooltip(d, x, y) {
-      pointEls.forEach(p => {
-        if (p.data === d) {
-          p.el.style.opacity = '1';
-          p.el.style.filter = 'drop-shadow(0 2px 8px rgba(0,0,0,0.25))';
-        } else {
-          p.el.style.opacity = '0.15';
-          p.el.style.filter = 'none';
-        }
-      });
-      paretoLine.style.opacity = '0.15';
-      const rect = chartWrap.getBoundingClientRect();
-      const svgRect = svg.getBoundingClientRect();
-      const scaleX = svgRect.width / W;
-      const scaleY = svgRect.height / H;
-      const tipX = x * scaleX + (svgRect.left - rect.left);
-      const tipY = y * scaleY + (svgRect.top - rect.top);
-      const successColor = MODEL_COLORS[d.model] || '#333';
-      const agentName = AGENT_DISPLAY[d.agent] || d.agent.replace(/_/g, ' ');
-      const modelName = MODEL_DISPLAY[d.model] || d.model;
-      tooltip.innerHTML =
-        '<div class="tooltip-agent">' + agentName + '</div>' +
-        '<div class="tooltip-model">' + modelName + '</div>' +
-        '<div class="tooltip-stats">' +
-          '<div><div class="tooltip-stat-label">Success</div><div class="tooltip-stat-value" style="color:' + successColor + '">' + (d.success * 100).toFixed(0) + '%</div></div>' +
-          '<div><div class="tooltip-stat-label">Cost/task</div><div class="tooltip-stat-value">$' + d.cost.toFixed(2) + '</div></div>' +
-        '</div>';
-      tooltip.classList.add('visible');
-      const tipW = tooltip.offsetWidth;
-      const tipH = tooltip.offsetHeight;
-      let left = tipX + 18;
-      let top = tipY - tipH / 2;
-      if (left + tipW > rect.width - 8) left = tipX - tipW - 18;
-      if (top < 4) top = 4;
-      if (top + tipH > rect.height - 4) top = rect.height - tipH - 4;
-      tooltip.style.left = left + 'px';
-      tooltip.style.top = top + 'px';
-    }
-    function hideTooltip() {
-      applyFilters();
-      tooltip.classList.remove('visible');
-    }
-    pointEls.forEach(({ el: g, data: d, x, y }) => {
-      g.addEventListener('mouseenter', () => showTooltip(d, x, y));
-      g.addEventListener('mouseleave', hideTooltip);
-    });
-    legendItems.forEach(item => {
-      item.addEventListener('click', () => {
-        const type = item.dataset.filter;
-        const value = item.dataset.value;
-        activeFilters[type] = (activeFilters[type] === value) ? null : value;
-        applyFilters();
-      });
-    });
-  }
-  // Expose globally so main.js can re-render on theme toggle
-  window.renderParetoChart = renderParetoChart;
-})();

styles.css CHANGED Viewed

@@ -404,7 +404,9 @@ section{position:relative;z-index:1;padding:80px 0}
 }
 .filter-select:hover,.filter-select:focus{border-color:var(--cyan);outline:none;box-shadow:0 0 12px var(--cyan-dim)}
-.table-wrap{overflow-x:auto;border-radius:12px;border:1px solid rgba(255,255,255,0.1);background:rgba(10,10,10,0.8);box-shadow:0 2px 16px rgba(0,0,0,0.3)}
 .lb-table{width:100%;border-collapse:collapse;font-size:.78rem}
 .lb-table thead{background:rgba(255,255,255,0.06)}
 .lb-table th{

 }
 .filter-select:hover,.filter-select:focus{border-color:var(--cyan);outline:none;box-shadow:0 0 12px var(--cyan-dim)}
+.table-wrap{overflow-x:auto;overflow-y:auto;max-height:600px;border-radius:12px;border:1px solid rgba(255,255,255,0.1);background:rgba(10,10,10,0.8);box-shadow:0 2px 16px rgba(0,0,0,0.3);position:relative}
+.table-wrap.lb-can-scroll{-webkit-mask-image:linear-gradient(to bottom,#000 88%,transparent 100%);mask-image:linear-gradient(to bottom,#000 88%,transparent 100%)}
+.lb-table thead{position:sticky;top:0;z-index:1;backdrop-filter:blur(8px);-webkit-backdrop-filter:blur(8px)}
 .lb-table{width:100%;border-collapse:collapse;font-size:.78rem}
 .lb-table thead{background:rgba(255,255,255,0.06)}
 .lb-table th{