AI Engineer YouTube · May 30, 2026

How We Built Zeta2: Training an Edit Prediction Model in Production — Ben Kunkle, Zed

How We Built Zeta2: Training an Edit Prediction Model in Production — Ben Kunkle, Zed video thumbnail
Why it matters

To validate settled data, Zed ran 10 frontier model predictions per example and measured Levenshtein distance to the final state. For 100,000 training examples that is a million frontier model requests, which is prohibitively expensive. The fix: Zeta 2's student model now approaches teacher quality, so they run it 50 t

My takeaway: To validate settled data, Zed ran 10 frontier model predictions per example and measured Levenshtein distance to the final state. For 100,000 training examples that is a million frontier model requests, which is prohibitively expensive. The fix: Zeta 2's student model now approaches teacher quality, so they run it 50 t