Swirlsmith
Back to leaderboard
#4

Kilo Code

๐Ÿ”“ Open

VS Code extension forked from Cline/Roo. Multi-provider, MCP support, custom modes, diff-based editing with full terminal integration.

๐Ÿ’ฐ Open-source VS Code/JetBrains extension ยท vscode, jetbrains

74.3Overall Score
Non-Gameable Scoring

Scores are derived from established benchmarks, adjusted for harness-specific performance across four dimensions: Coding, Reasoning, Tool Use, and Autonomy.

Each dimension starts from public benchmark data and applies harness-specific modifiers based on tool integration, context handling, and orchestration quality. The overall score is a weighted composite that penalizes narrow optimization.

ModelOverall
Claude Opus 4.674.3
Gemini 3.1 Pro63.9
Kimi K2.562.5
GPT-5.458.0
Claude Sonnet 4.657.2
MiniMax-M2.755.3
DeepSeek R148.9
Qwen 3.547.7
Gemini 3 Pro44.6
MiMo-V2-Flash42.6