Benchmark RadarCODA-BENCH
CODA-Bench is a benchmark for evaluating AI agents on data-intensive analytical tasks in a Linux sandbox with 1,009 tasks across 31 communities, requiring data discovery, code generation, and correct answers.
- Released
- 2026-06-13
- Readiness
- Runnable
- Primary field
- General AI
Why it matters
Existing benchmarks evaluate code or data capabilities in isolation. CODA-Bench jointly evaluates both, reflecting real development scenarios with complex file systems and large-scale data, revealing gaps in agentic data intelligence.
Motivation
Advanced agents are increasingly demonstrating the potential to operate as autonomous engineers, creating a growing demand for evaluation benchmarks that capture the complexity of real-world development.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.