New research · 2026
Discoverable by Design
How API description quality drives AI agent performance
New research benchmarking 8,467 operations across 70 real enterprise APIs: improving API description quality and search lifts agent retrieval accuracy from 59% to 94% — better data beats a bigger model.
- 94%agent retrieval accuracy after improving descriptions and search
- ~80%model-cost reduction using a smaller model, with no loss in selection accuracy
- 43%of catalogued API operations have both a summary and a description
