AI Code Contribution Calculator
Work out what share of your merged PRs is AI-authored or AI-assisted, adjust it for quality, and see how AI code stacks up against human code.
Your repos / periods
These are your own numbers - illustrative, editable estimates. Quality scores are whatever your team measures (review rubric, revert rate, defect density), not a benchmark. AI-authored + AI-assisted are capped at total PRs; anything left over is treated as human-only.
Results
250 PRsAI-touched share scaled by AI vs human quality.
How much of your code is really AI?
"AI writes X% of our code" is an easy headline and a hard number. Counting AI-authored versus AI-assisted PRs separately is a start, but raw volume overstates value if the AI bucket ships lower-quality changes. This calculator splits contribution into AI-authored, AI-assisted and human, then discounts (or rewards) the AI share based on how its average quality compares to your human baseline - a quality-adjusted contribution that's harder to game.
The quality scores are yours to define. Feed them from a review rubric, revert and defect rates, or a lightweight estimate like our PR quality estimator. To see how contribution moves your delivery numbers, pair it with the DORA metrics calculator and the cycle time calculator.
For the reasoning behind treating AI and human code on the same bar, read our guides on AI vs human code quality and measuring AI coding tools.
Frequently asked questions
Measure what your AI actually ships
FeatureFactory helps you plan, build and measure AI web development - tracking AI contribution and quality across your whole repo instead of a spreadsheet guess. The feature factory that measures what it ships.
Get early access