Higher scoring coding models did not boost developer productivity
Higher-Scoring Coding Models Did Not Get More Work Done
Software EngineeringArtificial IntelligenceMachine Learning
Summary
People often assume that newer coding AI models that get better scores on tests help programmers get more done. The authors studied what happened when a top-scoring coding model was temporarily unavailable, forcing users to switch to a lower-scoring model. They looked at the number of code changes merged into projects before, during, and after this switch and found no clear improvement in productivity when the better model was available. This suggests that higher test scores do not necessarily translate to more work completed in real coding tasks.
What this means in practice
- •For software engineering managers: Make informed decisions about selecting AI coding assistants by considering real productivity impact rather than just benchmark scores.
- •For ai model developers: Prioritize developing coding models that improve actual developer output over those that only score better on benchmarks.
Authors
Yining Hua
Abstract
New coding models advertise themselves on score improvements of coding benchmarks. Yet no study has tested if a model that scores higher on these benchmarks gets more real work done. On June 12, 2026, a U.S. export control directive banned access to Claude Fable 5 for 19 days, which forced users to switch to other models. During this time, most former Fable 5 users switched to Opus 4.8 (11 points lower on SWE-Bench Pro compared to Fable 5). We treat this suspension as a natural experiment, where the people and projects stayed the same and only the model changed, so we could isolate the effect of Fable 5 on user productivity. We tracked GitHub pull requests of 297 Fable 5 users before, during, and after the suspension. We used 8,908 developers who stayed on Opus 4.8 all the time as the control group. We also repeated the comparison across five model upgrades before and after the suspension in 2026 (1,208 developers, 15,442 pull requests). Fitting a difference-in-differences model with developer and date fixed effects, we found that having access to Fable 5 increased the 7-day merge rate of pull requests by only 1.4 percentage points [-1.4; +4.3], which was not statistically significant. Similarly, no significant increases of merge rates were found with the upgrades (weighted mean: -0.9 percentage points [-1.9; +0.2]); all five estimates were negative (95% intervals included 0, did not differ from each other). We also did not observe a significant increase in the number of pull requests merged per week (+2.4% [-4.9%; +10.2%]).