What is OpenAI LifeVal? The everyday tasks benchmark
What is OpenAI LifeVal? The everyday tasks benchmark OpenAI LifeVal evaluates agents based on 169 everyday tasks: all published scores, the numbers per reasoning effort and what the benchmark does not show.
OpenAI has developed a benchmark called LifeVal, which evaluates AI agents based on 169 everyday tasks, such as enrolling a child in school, managing a household budget, or interacting with government services. The benchmark measures the average reward, a score between 0 and 1, with higher scores indicating better performance. GPT-6 Astra achieved a score of 80.6% on the benchmark at xhigh reasoning effort, while previous models, including GPT-5.2, GPT-5.4, GPT-5.5, and GPT-5.6, scored lower, with GPT-5.2 scoring 49.3%.
The benchmark also breaks down scores by reasoning effort, with GPT-6 Astra achieving scores ranging from 66.6% at low effort to 80.6% at xhigh effort.
Written by urgent.news from Dev.to's report — not a translation of it. Machine-written — may contain errors; check the original before relying on it.
