mteb-leaderboard - ML Model Leaderboard and Benchmark Query Guide
Guidance for querying ML model leaderboards and benchmarks including MTEB and HuggingFace
Tags
Updated: 2026-05-09Capabilities
Typical Inputs
What this skill does
- Find top-performing models on benchmarks
- Query current leaderboard standings
- Compare model performance across benchmarks
- Identify authoritative data sources
- Verify data temporal alignment
- Access live leaderboard data
- Validate model eligibility
- Cross-reference multiple sources
- Search for raw data endpoints
- Extract data from interactive pages
- Check data source timestamps
Inputs
- Benchmark name
- Model name
- Target date or timeframe
- Official leaderboard websites
- GitHub repositories
- API endpoints
- JSON data files
- CSV data exports
Outputs
- Model rankings
- Benchmark comparison results
- Top-performing model list
- Leaderboard standings
- Data source references
Requirements
- Internet access
- Web browsing capability
- HuggingFace Spaces access
