AI-assisted software engineering has seen the emergence of several benchmarks to measure the capabilities of LLMs. Android developers face specific challenges that aren't covered by existing benchmarks, so we created one that focuses on a north star of high quality Android development.
Model Score (%) info Average percentage of 100 test cases successfully resolved across 5 runs for each model
arrow_range Cl range (%) info Expected performance range, reflecting the results' statistical reliability (p-value < 0.05)
Avg latency (h) info Average time taken to solve 100 tasks across 5 runs
Avg cost ($) info Average cost per full benchmark run
Claude Opus 5 91.8
87.4 — 95.7 13.2 $175.4
Claude Fable 5 90.9
86.2 — 95.4 8.7 $144.3
GPT 5.6 Sol 90.8
85.6 — 95.6 14.2 $150.0